- Email: [email protected]
Oesophageal diverticula
Best Practice & Research Clinical Gastroenterology Vol. 18, No. 1, pp. 3–17, 2004 doi:10.1053/ybega.2004.430, available online at http://www.sciencedirect.com
1 Oesophageal diverticula Mario Costantini MD Giovanni Zaninotto*
MD, FACS
Professor
Christian Rizzetto MD Surendra Narne MD Ermanno Ancona MD, FACS Clinica Chirugica IV, Department of Medical and Surgical Sciences, University of Padua, 2, via Giustiniani, Padua, Italy
Oesophageal diverticula are rare. They are most commonly seen at the pharyngo-oesophageal junction (Zenker’s diverticula) or at the distal oesophagus (epiphrenic diverticula). In both cases they are caused by altered motility which results in abnormal intraluminal pressure and the pushing of the oesophageal mucosa through focal weaknesses of the muscular wall (pulsion diverticula). The established surgical treatment for these diverticula therefore consists of eliminating the functional obstruction causing the disease (myotomy), associated with resection of the diverticulum (diverticulectomy) or its suspension (diverticulopexy). Recently, the spread of minimally invasive surgery has also led the application of such techniques to the treatment of oesophageal diverticula. Endoscopic diverticulostomy with stapler, laser or coagulation, through a rigid or flexible endoscope, has been demonstrated to be a valid treatment for Zenker’s diverticula—as an alternative to surgery—especially in high-risk patients. On the other hand, laparoscopic treatment of epiphrenic diverticula has recently been introduced with encouraging results. However, because the disease is rare, more experience is required in order to allow definitive conclusions. Key words: oesophageal diverticula; Zenker’s diverticulum; surgery; minimally invasive surgery; endoscopy; laparoscopy.
Oesophageal diverticula are rare. They can occur at any oesophageal level, but their aetiology, symptoms and therapeutic requirements suggest a classification in three categories: pharyngo –oesophageal, parabronchial and epiphrenic diverticula. Moreover, the classical Rokitansky classification1 still provides useful information on the aetiopathogenesis of oesophageal diverticula. Traction diverticula are the result of a chronic inflammatory process starting from the mediastinal lymph nodes (usually from a granulomatous disease) which involves the oesophageal wall. These diverticula are commonly seen in the vicinity of the carina ( parabronchial diverticula), they tend to be * Corresponding author. Tel.: þ39-49-821-1718; Fax: þ39-49-821-3152. E-mail address: [email protected] (G. Zaninotto). 1521-6918/04/$ - see front matter Q 2003 Elsevier Ltd. All rights reserved.
4 M. Costantini et al
small with a large neck, and they rarely become symptomatic or require medical attention. On the other hand, pulsion diverticula are the result of an altered pressure gradient inside the oesophageal lumen, which determines their formation through loci minoris resistentiae in the oesophageal wall. These can be naturally present, as the Killian’s triangle where the pharyngo-oesophageal diverticula develop, or due to tearing of the oesophageal wall above a zone of altered motility, as in the case of epiphrenic diverticula. The treatment for pulsion diverticula has traditionally been performed using an open surgical approach (cervicotomy or thoracotomy), but recent developments in the socalled ‘minimally invasive’ procedures have led to the application of such techniques in this field too. In this chapter, we aim to review the most recent acquisitions in the video-endoscopic treatment of oesophageal diverticula, describing the surgical technique and discussing our own and other authors results, comparing them particularly with the traditional open surgical approach.
PHARYNGO-OESOPHAGEAL (ZENKER’S) DIVERTICULA Pharyngo-oesophageal diverticula are protrusions of pharyngeal mucosa through a weak zone in the posterior wall of the pharynx, limited inferiorly by the upper border of the cricopharyngeal muscle and laterally by the oblique fibres of the thyropharyngeal muscle, the so-called Killian’s triangle2 (Figure 1(A)). Although described for the first time by Ludlow3, the cricopharyngeal pouch is better known by the name of a German
Figure 1. (A) Radiological image of a pharyngo-oesophageal (Zenker’s) diverticulum. (B) The same patient, after myotomy of the UOS and diverticulectomy.
Oesophageal diverticula 5
pathologist, Frederick Albert von Zenker who, together with von Ziemssen, published a review of 27 patients with this disease.4 Although not the first to describe this condition, Zenker takes merit for realising that the pathogenesis of the diverticula lies in an increased intra-pharyngeal pressure. Nearly a century later, the role of the upper oesophageal sphincter (UOS) in causing excessive intra-pharyngeal pressure during swallowing was demonstrated by manometric and cineradiographic studies5 and, from then on, the standard surgical treatment for Zenker’s diverticulum consisted in myotomy of the UOS and resection or pexis of the pouch, or even myotomy alone for small diverticula.6 Alternative endoscopic procedures that divided the septum between the diverticula and the oesophageal wall using a cautery or laser7,8 were also described, but gained little popularity because of the high risk of severe complications. The situation changed in 1993, when Collard proposed simultaneously dividing and suturing the diverticular and oesophageal wall using a laparoscopic stapler introduced through a special endoscope (the Weerda diverticuloscope). With this technique, the anterior wall of the diverticulum and the posterior wall of the oesophagus were sealed with a double row of staples along the cut edges, thus preventing leakage, mediastinitis or bleeding.9 This procedure rapidly became widespread and is now often considered the treatment of choice for Zenker’s diverticula.10,11 In 1995, moreover, Ishioka et al in Brazil12 and Mulder et al in the Netherlands13 reported their first results from using a flexible endoscope for cutting the diverticular septum.
Endoscopic stapling diverticulostomy The operation is performed under general anaesthesia with endotracheal intubation. The patient is placed supine on the operating table, with a small pillow under the upper torso and the head hyperextended. The surgeon sits behind the patient’s head. A Weerda diverticuloscope (Karl Storz, Tuttingen, Germany) is inserted in the hypopharynx, positioning its anterior blade in the oesophageal lumen and the posterior blade in the diverticulum. A 5 mm diameter telescope is passed through the scope. After visualization of the septum between the oesophagus and the diverticulum, the diverticuloscope is fixed with the help of a chest support. The length of the diverticulum is accurately measured with a graduated rod. A disposable surgical endostapler (EndoGIA 30, United States Surgical Corp. Norwalk, CT) is then inserted through the Weerda scope to divide the septum between the diverticulum and the oesophageal lumen and to perform the diverticulo – oesophagostomy. The device has been modified by shortening its anvil, thus enabling complete tissue stapling and sectioning down to the very bottom of the diverticulum. The anvil is placed in the lumen of the diverticulum and the cartridge in the oesophageal lumen. Stapling sutures the posterior oesophageal wall to the wall of the diverticulum over a length of about 30 mm and the tissue coming between three rows of staples on each side is transected (Figure 2). A second stapling step may be required, depending on the actual size of the diverticulum. Electrocoagulation with endosurgical scissors may be used to complete the dissection of the septum at the distal end on the stapled line. The suture lines are then checked for haemostasis and the scope is removed. The procedure takes about 20 minutes. After radiological assessment with a hydrosoluble contrast the following day, the patient starts a liquid diet and is usually discharged after two days. The diet is restored to normal in about a week.
6 M. Costantini et al
Figure 2. Schematic representation of the transoral stapling diverticulostomy. The diverticuloscope is inserted through the mouth, positioning its anterior blade in the oesophageal lumen and the posterior blade in the diverticulum. With a modified surgical endostapler the septum between the diverticulum and the oesophageal lumen is divided and the diverticulo-oesophagostomy performed.
Fibero-optic endoscopic treatment Other endoscopic techniques have been employed to divide the septum between the oesophageal lumen and the diverticulum, using a flexible endoscope. These can be valuable in elderly patients at high risk for general anaesthesia or with contraindications for the previously-described techniques (e.g. the inability to open their mouth wide, or diverticula less than 2 cm in size). Initially this approach was reserved for patients at high surgical risk, but it has progressively been extended, in some centres at least, to all patients referred for treatment. Patients are under conscious sedation or propofol with the assistance of an anaesthesiologist. The procedure is usually performed in the endoscopy unit. It is generally possible to obtain a good view of the septum: a nasogastric tube is useful to optimize the exposure and it helps to protect the anterior oesophageal wall. The incision is made starting from the rim between the oesophageal orifice and the opening of the diverticulum. This can be done using electrocautery, with a combination of cutting and coagulation currents.12,13 Other authors use the argon plasma coagulator (APC)14 or the CO2-laser.8,15 The aim of the procedure is to obliterate the septum completely, thus achieving a wide opening between the diverticulum and the oesophagus (diverticulostomy). Only one session is usually needed for small diverticula ð# 2 cmÞ; whereas repeated sessions become necessary if
Oesophageal diverticula 7
they are larger. At any one session, a 1.5 – 2.0 cm incision is performed, to be repeated in 1 week. A transparent hood attached to the tip of the endoscope is useful to keep the scope away from the oesophageal wall.16 A naso-gastric tube is usually placed for nutritional purposes for two days. Results and discussion All authors agree that a fundamental step in therapy for Zenker’s diverticula lies in dividing the crico-pharyngeous muscle fibres (i.e. the upper oesophageal sphincter). Obvious though it may seem today, this notion is quite a recent acquisition, based empirically on the finding of much lower complication (i.e. leakage) and recurrence rates when miotomy is added to the simple diverticulectomy.6 Definitive support for this convinction came from recent elegant studies with combined manometry and videofluorography5 which demonstrated that the muscle pathology of the UOS with inflammation and fibrosis restricts the opening of the UOS, leading to a higher pressure of the bolus arriving in the hypopharynx and ultimately to the formation of a pulsion diverticulum. Surgical crico-pharyngeal myotomy, alone in the case of small ð, 2 cmÞ diverticula, or combined with either diverticulectomy or diverticulum suspension (diverticulopexy) ensures symptom relief in nearly all treated patients17 (Figure 1(B)). The related morbidity may involve some local haematomas and recurrent nerve palsy, in addition to leakages—which occur in about 2% of cases.6 It should be emphasized, however, that the mortality rate of this procedure in frail, elderly patients, is far from negligible, mostly due to cardiopulmonary complications. The division of the UOS muscle fibres can also be achieved by endoscopic techniques, however, by dividing the septum between the oesophageal lumen and the diverticulum using laser, cautery or stapling devices. The results listed in Table 1 show that satisfactory results can be achieved in about 90% of patients with the endoscopic approach too. This figure is very attractive as the procedure demands only about 20 minutes of general anaesthesia or, better still, it can be done under conscious sedation in the endoscopic unit. This is particularly relevant, given the intrinsic characteristics of Table 1. Results of endoscopic treatment of Zenker’s diverticula.
Author Ishioka et al (1995)12 Scher, 199618 Peracchia et al (1998)10 Baldwin and Toma (1998)19 Mulder (1999)14 Hashiba et al (1999)20 Stoecki, (2001)21 Mattinger and Ho¨rmann (2002)22 Counter et al (2002)23 a
Number of patients 42 36 92 51 167 47 30 52 31
Method
Morbidity (and mortality)
Follow-up (months)
Cautery Stapling Stapling Stapling APC laser Cautery Stapling CO2 laser Stapling
2 5a None 1 leak 12 (1)b 1 bleeding 2 mucosal tears 6 (1)c 3 leaks
38 9.3 23 15 ? ? 13.2 ? 60
Good results (%) 92.8 88.9 94.6 100 100 95.8 93.3 84.6 78.0
One perforation, one post-op fever, one transient recurrent nerve palsy, two dental injuries. Seven mediastinal emphysema, four bleeding; one death for pulmonary embolism. c Three haemorrhages, one recurrent nerve palsy, two mediastinitis; one death from myocardial infarction. b
8 M. Costantini et al
these patients. The risks of complications and even death cannot be completely overcome by these techniques, but they are lower than with open surgery, and the endoscopic treatment of Zenker’s diverticula, with its ‘minimally invasive’ nature, spread very rapidly among most surgeons, and is considered as the treatment of choice for this condition at most centres.10,11 So far there have been no prospective trials comparing the different endoscopic treatment options with surgery. Information comes from retrospective series or prospectively recorded case series, using one or more of these techniques, as in the case of all the series listed in Table 1. Moreover, given the rarity of the disease and its prevalence in elderly patients, often with severe co-morbidities, it is unlikely that randomized studies can be performed in the near future. Our recent experience consists of 24 patients treated with endoscopic stapling diverticulostomy, and 35 patients treated with surgical myotomy. Diverticulectomy was added in 21 of the latter, whereas the diverticulum was inverted below the pharyngeal muscle layer in eight patients who had a diverticulum of about 2 cm. In five patients only UOS myotomy was performed because their diverticulum was only about 1 cm. These two groups of patients have recently been reviewed.24 The choice of the operation was prompted by the size of the diverticulum and the surgical risk: patients with a high surgical risk and diverticula $ 3 cm and # 5 cm were advised to have the endoscopic procedure, whereas open surgery was recommended to younger patients at low risk, or with diverticula , 3 cm or . 5 cm. We have only sporadic experience with flexible endoscopic treatment, so the few patients concerned are not considered for the present analysis. No mortality was observed in the two groups of patients. The overall morbidity was 8.6% and all the complications were observed in the open-surgery group: there were two leakages, treated conservatively, two haemorrhages with cervical haematoma requiring drainage, and one pericarditis. The post-operative hospital stay was significantly shorter in the endoscopic group. The median duration of follow-up was 41 (range 1 –101) months. Three patients died during the follow-up for unrelated causes and two were lost to follow-up, so an adequate follow-up was obtained in 53 patients, 20 in the endoscopic group and 33 in the open surgery group. After the operation there was a statistically significant improvement in the symptom score in both groups. In the endoscopic group, however, three patients still complained of severe dysphagia and were considered as procedure failures (15%). No patients complained of severe dysphagia in the open-surgery group ðP , 0:05Þ: A posterior pouch was still evident (Figure 3) in all patients treated endoscopically and studied with post-operative barium swallow; most of the patients were asymptomatic, however. Similar results are reported by the Brussels group25 which first proposed endoscopic stapling for Zenker’s diverticula (Table 2). They retrospectively compared their experience with endoscopic treatment with both the stapler (25 patients) and the laser (54 patients), against traditional myotomy alone (seven patients), plus resection (12 patients) or pexis (36 patients), and against diverticulectomy alone (29 patients). Patients treated endoscopically had a shorter hospital stay and fasting period after the operation, as well as fewer complications (although two cervical abscesses and mediastinitis were recorded). Symptom outcome was less favourable than with open surgery, however, and only 57% of patients with a diverticulum , 3 cm were satisfied with the treatment, whereas this was true of 98% of patients treated with open surgery ðP , 0:05Þ: This difference was less evident in patients with larger diverticula (. 3 cm), however: 80 versus 96%. But still, in the series as a whole, only 75% of patients treated
Oesophageal diverticula 9
Figure 3. Barium swallow after endoscopic stapling diverticulostomy. A posterior pouch is still visible in all the patients, representing the diverticulum widely anastomosed to the oesophageal lumen.
endoscopically were symptom-free at the follow up, as opposed to 97% of the patients who had open surgery. There is only one other paper comparing endoscopic stapling diverticulostomy with open surgery in two small groups of patients (eight patients each)26; this paper emphasizes the significantly shorter operation time, post-operative hospital stay and time to oral intake, and the lower hospital charges for the patients treated
10 M. Costantini et al
Table 2. Retrospective studies comparing surgical myotomy and endoscopic treatment for Zenker’s diverticula. Surgical myotomy
Author Smith, 200226 Gutshow, 200225 Zaninotto, 200324 a b c d
Number of patients 8 67 34
Complications – 6b 5d
Endoscopic treatment
Good results (%) 100 97 100
Number of patients 8 86 24
Complications
Good results (%)
1a 3c -
100 75 87.5
p , 0:05 p , 0:05
One bleeding. Five leakages; one death for myocardial infarction. Two cervical abscesses and mediastinitis; one dental injury. Two leakages, two haematomas, one pericarditis.
endoscopically. Symptomatic relief was reportedly obtained in all patients in both groups, with only one minor complication (bleeding) in a patient of the endoscopic group. However, no details are given of symptom assessment or length of follow-up. It is difficult to draw final conclusions from these studies. Endoscopic treatment is very attractive because it is less invasive and has a lower complication rate although, when compared to surgery, it is sometimes less effective in relieving dysphagia completely. Endoscopic stapling also has other drawbacks, mainly related to the size of the diverticulum: in the case of a small diverticulum ð, 2 cmÞ the stapler anvil is too long to be properly accommodated inside the pouch and the cricopharyngeal fibres cannot be transected completely. In this sense, diverticulostomy with the laser or cautery may be more effective. On the other hand, very large diverticula ð. 5 cmÞ plunging into the mediastinum carry the risk of vascular lesions if they are transected blindly. Examples of minor drawbacks include patients who cannot open the mouth or those who cannot hyperextend the head. We observed a few such patients, where a surgical approach or the flexible endoscopic technique was used. In conclusion, Zenker’s diverticula can be effectively treated with either endoscopic diverticulostomy or open surgery. Both approaches have advantages and disadvantages. An individual approach to Zenker’s diverticula should therefore be recommended: high-risk patients with medium-size diverticula are probably better served by diverticulostomy. Open surgery should be recommended for small ð, 2 cmÞ or giant Zenker’s diverticula or in patients with a low surgical risk.
EPIPHRENIC DIVERTICULA These lesions occur less frequently than Zenker’s diverticula, with a ratio of 1:527, although this lower prevalence may be due to the fact that only a minority of these diverticula are symptomatic (about 25%). They are usually single, but it is not infrequent to find two or more synchronous diverticula (in up to 25% of cases, in our experience) as the result of the complex motor disorder behind their pathogenesis. Symptoms may vary greatly from one patient to another. Symptoms may be absent, the diverticulum
Oesophageal diverticula 11
being discovered by imaging studies performed for other reasons. When present, the most frequent symptoms are dysphagia and regurgitation of indigested food; these occur in up to 50% of the patients seeking medical attention. Retrosternal pain or heartburn, mimicking gastro-oesophageal reflux disease (GORD), may also be reported. In other instances, weight loss, halitosis, coughing and recurrent pneumonia are the presenting signs. Complications (ulcerations, bleeding) are rare these days, but spontaneous perforation (with a Boerhaave-like mechanism) may occasionally be seen. The diagnosis is radiological (Figure 4(A)): the spherical protrusion of the diverticulum from the oesophageal contour is easily identifiable. Tertiary contractions can often be seen, but the altered oesophageal motility is better documented at manometry. It should be noted, however, that standard manometry fails to document oesophageal dysmotility in about one in four patients. Prolonged 24-hour motility monitoring may help in discovering minor abnormalities.28 The treatment for these diverticula is surgical. However, the mere presence of an epiphrenic diverticulum is not, per se, an indication for surgery. The decision whether to operate must be balanced between the patient’s symptoms, the complication and operative risk, and the surgical expertise locally available, considering the rarity of the disease. A ‘masterful inactivity’, as described by Orringer29, in asymptomatic or mildly
Figure 4. (A) Barium swallow of a patient with an epiphrenic diverticulum. (B) The same patient, after laparoscopic diverticulectomy, oesophageal myotomy and Dor’s antireflux fundoplication.
12 M. Costantini et al
disturbing diverticula is good practice even if, in times of endoscopic staplers and devices, an oesophageal diverticulum may represent a tempting trophy for a hyperactive surgeon. The traditional surgical treatment for an epiphrenic diverticulum consists of an oesophageal myotomy, the diverticulectomy (or diverticulopexy), and an anti-reflux procedure, usually performed through a thoracotomy. More recently, developments in endoscopic surgical techniques have enabled the application of such techniques to the surgical treatment of diverticula of the lower oesophagus too. Thoracoscopic approach The first to be employed, as the logical consequence of the traditional thoracotomic route, was the thoracoscopic approach30 – 32, usually on the right. The patient is placed in the lateral decubitus position, and a double lumen tracheal tube is used because collapsing the lung facilitates the operation. Four ports are required: one for the camera, one for the lung retractor and two for the operating devices. With the help of intraoperative endoscopy, the diverticulum is identified and slowly isolated from the surrounding tissues with blunt dissection. Dissection is facilitated by using Babcock-like forceps applied to the diverticulum. One or two cartridges of staples are then applied to the base of the diverticulum, performing the diverticulectomy. This is a delicate step because endoscopic staplers are difficult to apply parallel to the oesophagus, at the base of the diverticulum, and the junction between two lines of staples may allow for postoperative leakage. The new articulated endostaplers may overcome this problem. An alternative approach is to perform a mini-thoracotomy, to introduce a normal linear stapler for better suture line control. The presence of the endoscope (or a 48 F bougie) inside the oesophagus prevents the lumen being narrowed excessively. The muscle layer overlying the suture line are approximated with some stitches. The oesophagus is then rotated to facilitate the execution of a myotomy on the opposite side. The proximal level of the myotomy should begin well above the level of the diverticulum neck, and continue to the oesophagogastric junction, that is identified with the help of the endoscope. Air insufflated through the scope may identify any perforation of the mucosa overlooked in performing the myotomy (hydro-pneumatic test). A fundoplication is not usually performed, because it is very difficult to perform via this approach, and may be not considered strictly necessary, given the minimal dissection performed at the oesophago-gastric junction, and the potent anti-reflux drugs available today. Laparoscopic approach More recently, increasing confidence with the laparoscopic approach to the hiatal region for the treatment of GORD and achalasia, and the accessibility of the lower thoracic oesophagus through the hiatus, have prompted the use of the laparoscopic approach for the surgical treatment of epiphrenic diverticula.33 The position of the patient and surgeon, and the trocars sites are the same as for the laparoscopic treatment of functional diseases of the esophagogastric junction. The phreno – oesophageal membrane is divided, and the diaphragmatic crura exposed. The oesophagus is then isolated and encircled with a silicon tube for safe traction. Blunt dissection of the lower mediastinum is then begun, paying attention to remain on the oesophageal wall, until the diverticular pouch is identified and carefully isolated up to the superior margin of its neck. Peri-operative endoscopy is very helpful in this step. When the diverticulum is completely isolated, the diverticular neck is sutured and resected with a linear endoscopic stapler, with the endoscope or a bougie inside the oesophageal lumen to
Oesophageal diverticula 13
Figure 5. Laparoscopic diverticulectomy: after complete mobilization of the diverticulum, diverticulectomy is performed by firing one or two endostapler cartridges at the base of the diverticulum. The endoscope or a bougie are inserted into the oesophagus, to avoid narrowing of the lumen. With this approach, the instrument is parallel to the oesophageal axis. After the diverticulectomy, the procedure is completed by a myotomy (on the opposite side of the oesophagus) and a partial antireflux procedure.
avoid narrowing the lumen (Figure 5). The resected diverticulum is then removed through a port, and the sutured line is protected by over sewing the muscular edges with some interrupted or continued suture. A hydropneumatic test can be performed by inflating air through the scope in the oesophageal lumen to identify any leakages. The myotomy is then performed on the opposite side of the oesophagus, from a level well above the upper margin of the diverticulum and down to 1 –2 cm below the cardia, on the gastric side. The operation is then completed with a partial anterior fundoplication (following Dor’s technique). These last two steps are identical to those of laparoscopic treatment for oesophageal achalasia. Results and discussion Epiphrenic diverticula are rare. They are often asymptomatic and do not always require surgical treatment. A group of asymptomatic patients was followed up at the Mayo Clinic for a median of 9 years and none became symptomatic during said time.27 Other authors28 favour the treatment of such patients, based on the finding of sometimes lifethreatening pulmonary complications in 25– 45% of the cases they followed. However, the success rate with the traditional thoracotomic approach is only about 85% and the operation carries a considerable morbidity (namely leakages from the suture line) and, above all, an undeniable mortality. The basic principles of surgical treatment for
14 M. Costantini et al
epiphrenic diverticula were defined back in the 1960s and have changed little since. Areas of controversy still exist34, but most surgeons agree that complete relief of any distal oesophageal obstruction (with myotomy) is the mainstay of the treatment, even if a motor disorder cannot be found with function studies. The introduction of minimally invasive surgery certainly revitalized interest in this pathology, as in the case of several other surgical fields. A couple of successful case reports of thoracoscopic resection of epiphrenic diverticula, with31 or without32 myotomy, have recently been reported. Another study, using thoracoscopic diverticulectomy, alone or associated with endoscopic pneumatic dilation of the cardia in the case of evident motor abnormalities, reported a high incidence of post-operative leakages, prompting the authors to abandon such an approach.30 We have personally treated only one patient with this approach, performing a simple thoracoscopic diverticulectomy after forceful pneumatic dilation of the cardia. In this particular patient, we observed a radiological recurrence of the diverticulum, albeit asymptomatic, a year later. These experiences confirm the importance of oesophageal myotomy in treating such complex disorders and the need to avoid modifying a consolidated surgical technique simply to render it more suitable for a novel approach. A completely laparoscopic approach was proposed by Rosati et al in 199833, and in 2001 the same group reported on their whole experience in 11 patients treated in this way, with a median follow-up of 3 years.35 Their results were good in all the patients treated, with only one post-operative leakage that required a thoracotomy to be repaired. Other reports with this technique have been published, reporting the successful laparoscopic treatment of epiphrenic diverticula36,37, even after an iatrogenic perforation.38 Our personal experience consists of eight cases (Table 3). Diverticulectomy (with one or two endostapler cartridges), extramucosal myotomy on the opposite side of the oesophagus and anterior fundoplication were completed laparoscopically in all the patients. Morbidity was high, however, with three postoperative leakages, requiring prolonged naso-gastric suction, parenteral nutrition and antibiotics. All three patients healed with this conservative treatment. After a median follow-up of 36.5 months, all the eight patients are satisfied with the treatment and are eating normally, but persistent reflux symptoms make continuous Proton Pump Inhibitors (PPI) treatment necessary in two of them. For the time being, the relatively recent introduction of such techniques—and the rarity of the disease—do not provide sufficient experience to enable us to draw definitive conclusions, such as those which can now be drawn, for example, for the laparoscopic Table 3. Personal experience with the laparoscopic treatment of epiphrenic diverticula. Patient number
Sex
Age
Size
Manometry
Morbidity
Follow-up (months)
Result
1 2 3 4 5 6 7 8
M F F F M M M F
64 70 59 41 54 55 67 41
4 cm 4 cm 5 cm 4 cm 7 cm 6 cm 5 cm 7 cm
LOS and body dysmotility LOS and body dysmotility Body dysmotility Normal Body dysmotility Body dysmotility Achalasia Normal
– – Pleural effusion – Leakage Leakage – Leakage
51 45 40 40 33 29 24 18
Good Good Good Good GORD Good Good GORD
Oesophageal diverticula 15
treatment of GORD or achalasia. Some considerations are nonetheless worth stating. A laparoscopic approach to these diverticula is certainly feasible. It offers a good access to the mediastinal oesophagus and therefore makes all three steps of the consolidated surgical procedure (myotomy, diverticulectomy and fundoplication) feasible in a single session. It may therefore become the modern treatment of choice for these diverticula. Among its advantages, it is worth emphasizing the fact that the endoscopic stapler lies parallel to the axis of the oesophagus (and therefore also to the neck of the diverticulum). Further, the optimal visualization of the hiatal region facilitates myotomy on both the oesophageal and gastric sides, and makes anterior fundoplication easy. Further experience is needed with larger numbers of patients to confirm the validity of this approach, but the rarity of the disease, the relative difficulty of the treatment and its considerable, potentially severe morbidity make this treatment suitable only for specialized referral centres.
Practice points † Zenker’s and epiphrenic diverticula of the oesophagus are caused by a disordered oesophageal motility † surgical treatment of oesophageal diverticula requires the elimination of functional obstruction by means of myotomy † myotomy of the UOS can effectively be achieved by means of endoscopic techniques † minimally invasive surgery has been successfully employed for the treatment of oesophageal epiphrenic diverticula † in this context, laparoscopic access seems more promising than thoracoscopic access † because of the rarity of the disease, laparoscopic treatment of epiphrenic diverticula must be reserved for dedicated centres with expertise in laparoscopic surgery of the gastro-oesophageal junction and a good knowledge of oesophageal pathophysiology
Research agenda † further studies comparing endoscopic treatment of Zenker’s diverticula to open surgery are needed in order to clarify the indications of this technique † the laparoscopic treatment of epiphrenic oesophageal diverticula requires further evaluation in order to verify its efficacy and safety. Until more experience is acquired, this approach should not be considered the gold standard of therapy for such diverticula
REFERENCES 1. Posthletwait RW. Diverticula of the esophagus. Surgery of the Esophagus. Norwalk: Appleton Century Crofts, 1986, pp 129 –160. 2. Killian G. The mouth of the esophagus. Laryngoscope 1907; 17: 421– 428. 3. Ludlow A. A case of obstructed deglutition from a preternatural dilatation of, and a bag formed, in the pharynx. Medical Observations and Inquiries of the Society of Physicians 1760; 3: 85– 101.
16 M. Costantini et al 4. Zenker FA & von Ziemssen H. Krankheiten des Oesophagus. In von Ziemsen H (ed.) Handbuch der Speciellen Pathologie und Therapie, 7. FC Vogel: Leipzig, 1877. * 5. Cook IJ, Gabb M & Panagopoulos V. Zenker’s diverticulum is a disorder of upper esophageal sphincter opening. Gastroenterology 1993; 103: 1229–1235. 6. Payne WS. The treatment of pahryngoesophageal diverticulum: the simple and the complex. Hepatogastroenterology 1992; 39: 109 –114. 7. Dohlman G & Mattson O. The endoscopic operation for hypopharyngeal diverticula. Archives of Otolaryngology 1960; 71: 744–752. 8. Knegt PP, de Jong PC & van der Schans EJ. Endoscopic treatment of the hypopharyngeal diverticulum with CO2 laser. Endoscopy 1985; 17: 205–206. * 9. Collard JM, Otte IB & Kestens PJ. Endoscopic stapling technique of esophagodiverticulostomy for Zenker’s diverticulum. Annals of Thoracic Surgery 1993; 56: 573–576. * 10. Peracchia A, Bonavina L, Narne S et al. Minimally invasive surgery for Zenker’s diverticulum. Analysis of results in 95 consecutive patients. Archives of Surgery 1998; 133: 695–700. 11. Omote K, Feussner H, Stein HJ et al. Endoscopic stapling diverticulostomy for Zenker’s diverticulum. Surgical Endoscopy 1999; 13: 535–538. 12. Ishioka S, Sakai P, Maluf FF & Melo JM. Endoscopic incision of Zenker’s diverticula. Endoscopy 1995; 27: 433–437. 13. Mulder CJJ, Robijn RJ & Thies JE. Flexible endoscopic treatment of Zenker’s diverticulum: a new approach. Endoscopy 1995; 27: 438–442. 14. Mulder CJJ. Zapping Zenker’s diverticulum: gastroscopic treatment. Canadian Journal of Gastroenterology 1999; 13: 405–407. 15. van Overbeek J, Hoeksema PE & Edens ET. Microendoscopic surgery of the hypopharyngeal diverticulum using electrocoagulation or carbon dioxide laser. Annals of Otology, Rhinology and Laryngology 1984; 93: 34–36. * 16. Mulder CJJ, Costamagna G & Sakai P. Zenker’s diverticulum: treatment using a flexible endoscope. Endoscopy 2001; 33: 991–997. * 17. Sideris L, Qi Chen L, Ferraro P & Duranceau AC. The treatment of Zenker’s diverticula: a review. Seminars in Thoracic and Cardiovascular Surgery 1999; 11: 337 –351. 18. Scher RI & Richtsmeier WJ. Long-term experience with endoscopic stapled-assisted esophagodiverticulostomy for Zenker’s diverticulum. Laryngoscope 1998; 108: 200– 205. 19. Baldwin DL & Toma AG. Endoscopic stapled divertciulostomy: a real advance in the treatment of hypopharyngeal diverticulum. Clinical Otolaryngology 1998; 23: 244–247. 20. Hashiba K, de Paula AL, da Silva JG et al. Endoscopic treatment of Zenker’s diverticulum. Gastrointestinal Endoscopy 1999; 49: 93–97. 21. Stoecki SJ & Schmid S. Endoscopic stapled-assisted diverticulo-esophagostomy for Zenker’s diverticulum: patient satisfaction and subjective relief of symptoms. Surgery 2002; 131: 158–162. 22. Mattinger C & Ho¨rmann K. Endoscopic diverticulostomy of Zenker’s diverticulum: management and complications. Dysphagia 2002; 17: 34 –39. 23. Counter PR, Hilton ML & Baldwin DL. Long-term follow-up of endoscopic stapled diverticulotomy. Annals of the Royal College of Surgeons of England 2002; 84: 89–92. * 24. Zaninotto G, Narne S, Costantini M et al. Tailored approach to Zenker’s diverticula. Surgical Endoscopy 2003; 17: 129 –133. * 25. Gutschow CA, Hamoir M, Rombaux PH et al. Management of pharyngoesophageal (Zenker’s) diverticulum: which technique? Annals of Thoracic Surgery 2002; 74: 1677– 1683. 26. Smith SR, Genden EM & Urken ML. Endoscopic stapling technique for the treatment of Zenker diverticulum vs standard open-neck technique. Archives of Otolaryngology–Head and Neck Surgery 2002; 128: 141 –144. * 27. Benacci JC, Deschamps C, Trastek VF et al. Epiphrenic diverticulum: results of surgical treatment. Annals of Thoracic Surgery 1993; 55: 1109–1113. * 28. Nehra D, Lord RV, DeMeester TR et al. Physiologic basis for the treatment of epiphrenic diverticulum. Annals of Surgery 2002; 235: 346 –354. 29. Orringer MB. Epiphrenic diverticula: fact and fable. Annals of Thoracic Surgery 1993; 55: 1067– 1068. 30. Peracchia A, Bonavina L, Rosati R & Bona S. Thoracoscopic resection of epiphrenic esophageal diverticela. In Peters JH & DeMeester TR (eds) Minimally Invasive Surgery of the Foregut. St Louis: Quality Medical Publishing, 1994, pp 110– 116. 31. Saw EC, McDonald TP & Kam NT. Video-assisted thoracoscopic resection of an epiphrenic diverticulm with esophagomyotomy and partial fundoplication. Surgical Laparoscopy and Endoscopy 1998; 8: 145–148. 32. Stuart RC, Wyman A, Chan AW et al. Thoracoscopic resection of oesophageal diverticulum: a case report. Journal of the Royal College of Surgeons of Edinburgh 1996; 41: 118–119.
Oesophageal diverticula 17 * 33. Rosati R, Fumagalli U, Bona S et al. Diverticulectomy, myotomy and fundoplication through laparoscopy: a new option to treat epiphrenic esophageal diverticula? Annals of Surgery 1998; 227: 174–178. 34. Lerut T & Hiebert CA. Esophageal diverticula. In Pearson FG (ed.) Esophageal Surgery. New York: Churchill Livingstone, 2002, pp 507– 514. 35. Rosati R, Fumagalli U, Bona S et al. Laparoscopic treatment of epiphrenic diverticula. Journal of Laparoendoscopic and Advanced Surgical Techniques 2001; 11: 371–375. 36. Chami Z, Fabre JM, Navarro F & Domergue J. Abdominal laparoscopic approach for thoracic epiphrenic diverticulum. Surgical Endoscopy 1999; 13: 164–165. 37. Eubanks TR & Pellegrini CA. Minimally invasive treatment of esophageal diverticula. Seminars in Thoracic and Cardiovascular Surgery 1999; 11: 363 –367. 38. Tinoco RC, Tinoco AC & El-Kadre L. Perforated epiphrenic diverticulum treated by video laparoscopy. Surgical Endoscopy 1999; 13: 270–272.
1 Oesophageal diverticula Mario Costantini MD Giovanni Zaninotto*
MD, FACS
Professor
Christian Rizzetto MD Surendra Narne MD Ermanno Ancona MD, FACS Clinica Chirugica IV, Department of Medical and Surgical Sciences, University of Padua, 2, via Giustiniani, Padua, Italy
Oesophageal diverticula are rare. They are most commonly seen at the pharyngo-oesophageal junction (Zenker’s diverticula) or at the distal oesophagus (epiphrenic diverticula). In both cases they are caused by altered motility which results in abnormal intraluminal pressure and the pushing of the oesophageal mucosa through focal weaknesses of the muscular wall (pulsion diverticula). The established surgical treatment for these diverticula therefore consists of eliminating the functional obstruction causing the disease (myotomy), associated with resection of the diverticulum (diverticulectomy) or its suspension (diverticulopexy). Recently, the spread of minimally invasive surgery has also led the application of such techniques to the treatment of oesophageal diverticula. Endoscopic diverticulostomy with stapler, laser or coagulation, through a rigid or flexible endoscope, has been demonstrated to be a valid treatment for Zenker’s diverticula—as an alternative to surgery—especially in high-risk patients. On the other hand, laparoscopic treatment of epiphrenic diverticula has recently been introduced with encouraging results. However, because the disease is rare, more experience is required in order to allow definitive conclusions. Key words: oesophageal diverticula; Zenker’s diverticulum; surgery; minimally invasive surgery; endoscopy; laparoscopy.
Oesophageal diverticula are rare. They can occur at any oesophageal level, but their aetiology, symptoms and therapeutic requirements suggest a classification in three categories: pharyngo –oesophageal, parabronchial and epiphrenic diverticula. Moreover, the classical Rokitansky classification1 still provides useful information on the aetiopathogenesis of oesophageal diverticula. Traction diverticula are the result of a chronic inflammatory process starting from the mediastinal lymph nodes (usually from a granulomatous disease) which involves the oesophageal wall. These diverticula are commonly seen in the vicinity of the carina ( parabronchial diverticula), they tend to be * Corresponding author. Tel.: þ39-49-821-1718; Fax: þ39-49-821-3152. E-mail address: [email protected] (G. Zaninotto). 1521-6918/04/$ - see front matter Q 2003 Elsevier Ltd. All rights reserved.
4 M. Costantini et al
small with a large neck, and they rarely become symptomatic or require medical attention. On the other hand, pulsion diverticula are the result of an altered pressure gradient inside the oesophageal lumen, which determines their formation through loci minoris resistentiae in the oesophageal wall. These can be naturally present, as the Killian’s triangle where the pharyngo-oesophageal diverticula develop, or due to tearing of the oesophageal wall above a zone of altered motility, as in the case of epiphrenic diverticula. The treatment for pulsion diverticula has traditionally been performed using an open surgical approach (cervicotomy or thoracotomy), but recent developments in the socalled ‘minimally invasive’ procedures have led to the application of such techniques in this field too. In this chapter, we aim to review the most recent acquisitions in the video-endoscopic treatment of oesophageal diverticula, describing the surgical technique and discussing our own and other authors results, comparing them particularly with the traditional open surgical approach.
PHARYNGO-OESOPHAGEAL (ZENKER’S) DIVERTICULA Pharyngo-oesophageal diverticula are protrusions of pharyngeal mucosa through a weak zone in the posterior wall of the pharynx, limited inferiorly by the upper border of the cricopharyngeal muscle and laterally by the oblique fibres of the thyropharyngeal muscle, the so-called Killian’s triangle2 (Figure 1(A)). Although described for the first time by Ludlow3, the cricopharyngeal pouch is better known by the name of a German
Figure 1. (A) Radiological image of a pharyngo-oesophageal (Zenker’s) diverticulum. (B) The same patient, after myotomy of the UOS and diverticulectomy.
Oesophageal diverticula 5
pathologist, Frederick Albert von Zenker who, together with von Ziemssen, published a review of 27 patients with this disease.4 Although not the first to describe this condition, Zenker takes merit for realising that the pathogenesis of the diverticula lies in an increased intra-pharyngeal pressure. Nearly a century later, the role of the upper oesophageal sphincter (UOS) in causing excessive intra-pharyngeal pressure during swallowing was demonstrated by manometric and cineradiographic studies5 and, from then on, the standard surgical treatment for Zenker’s diverticulum consisted in myotomy of the UOS and resection or pexis of the pouch, or even myotomy alone for small diverticula.6 Alternative endoscopic procedures that divided the septum between the diverticula and the oesophageal wall using a cautery or laser7,8 were also described, but gained little popularity because of the high risk of severe complications. The situation changed in 1993, when Collard proposed simultaneously dividing and suturing the diverticular and oesophageal wall using a laparoscopic stapler introduced through a special endoscope (the Weerda diverticuloscope). With this technique, the anterior wall of the diverticulum and the posterior wall of the oesophagus were sealed with a double row of staples along the cut edges, thus preventing leakage, mediastinitis or bleeding.9 This procedure rapidly became widespread and is now often considered the treatment of choice for Zenker’s diverticula.10,11 In 1995, moreover, Ishioka et al in Brazil12 and Mulder et al in the Netherlands13 reported their first results from using a flexible endoscope for cutting the diverticular septum.
Endoscopic stapling diverticulostomy The operation is performed under general anaesthesia with endotracheal intubation. The patient is placed supine on the operating table, with a small pillow under the upper torso and the head hyperextended. The surgeon sits behind the patient’s head. A Weerda diverticuloscope (Karl Storz, Tuttingen, Germany) is inserted in the hypopharynx, positioning its anterior blade in the oesophageal lumen and the posterior blade in the diverticulum. A 5 mm diameter telescope is passed through the scope. After visualization of the septum between the oesophagus and the diverticulum, the diverticuloscope is fixed with the help of a chest support. The length of the diverticulum is accurately measured with a graduated rod. A disposable surgical endostapler (EndoGIA 30, United States Surgical Corp. Norwalk, CT) is then inserted through the Weerda scope to divide the septum between the diverticulum and the oesophageal lumen and to perform the diverticulo – oesophagostomy. The device has been modified by shortening its anvil, thus enabling complete tissue stapling and sectioning down to the very bottom of the diverticulum. The anvil is placed in the lumen of the diverticulum and the cartridge in the oesophageal lumen. Stapling sutures the posterior oesophageal wall to the wall of the diverticulum over a length of about 30 mm and the tissue coming between three rows of staples on each side is transected (Figure 2). A second stapling step may be required, depending on the actual size of the diverticulum. Electrocoagulation with endosurgical scissors may be used to complete the dissection of the septum at the distal end on the stapled line. The suture lines are then checked for haemostasis and the scope is removed. The procedure takes about 20 minutes. After radiological assessment with a hydrosoluble contrast the following day, the patient starts a liquid diet and is usually discharged after two days. The diet is restored to normal in about a week.
6 M. Costantini et al
Figure 2. Schematic representation of the transoral stapling diverticulostomy. The diverticuloscope is inserted through the mouth, positioning its anterior blade in the oesophageal lumen and the posterior blade in the diverticulum. With a modified surgical endostapler the septum between the diverticulum and the oesophageal lumen is divided and the diverticulo-oesophagostomy performed.
Fibero-optic endoscopic treatment Other endoscopic techniques have been employed to divide the septum between the oesophageal lumen and the diverticulum, using a flexible endoscope. These can be valuable in elderly patients at high risk for general anaesthesia or with contraindications for the previously-described techniques (e.g. the inability to open their mouth wide, or diverticula less than 2 cm in size). Initially this approach was reserved for patients at high surgical risk, but it has progressively been extended, in some centres at least, to all patients referred for treatment. Patients are under conscious sedation or propofol with the assistance of an anaesthesiologist. The procedure is usually performed in the endoscopy unit. It is generally possible to obtain a good view of the septum: a nasogastric tube is useful to optimize the exposure and it helps to protect the anterior oesophageal wall. The incision is made starting from the rim between the oesophageal orifice and the opening of the diverticulum. This can be done using electrocautery, with a combination of cutting and coagulation currents.12,13 Other authors use the argon plasma coagulator (APC)14 or the CO2-laser.8,15 The aim of the procedure is to obliterate the septum completely, thus achieving a wide opening between the diverticulum and the oesophagus (diverticulostomy). Only one session is usually needed for small diverticula ð# 2 cmÞ; whereas repeated sessions become necessary if
Oesophageal diverticula 7
they are larger. At any one session, a 1.5 – 2.0 cm incision is performed, to be repeated in 1 week. A transparent hood attached to the tip of the endoscope is useful to keep the scope away from the oesophageal wall.16 A naso-gastric tube is usually placed for nutritional purposes for two days. Results and discussion All authors agree that a fundamental step in therapy for Zenker’s diverticula lies in dividing the crico-pharyngeous muscle fibres (i.e. the upper oesophageal sphincter). Obvious though it may seem today, this notion is quite a recent acquisition, based empirically on the finding of much lower complication (i.e. leakage) and recurrence rates when miotomy is added to the simple diverticulectomy.6 Definitive support for this convinction came from recent elegant studies with combined manometry and videofluorography5 which demonstrated that the muscle pathology of the UOS with inflammation and fibrosis restricts the opening of the UOS, leading to a higher pressure of the bolus arriving in the hypopharynx and ultimately to the formation of a pulsion diverticulum. Surgical crico-pharyngeal myotomy, alone in the case of small ð, 2 cmÞ diverticula, or combined with either diverticulectomy or diverticulum suspension (diverticulopexy) ensures symptom relief in nearly all treated patients17 (Figure 1(B)). The related morbidity may involve some local haematomas and recurrent nerve palsy, in addition to leakages—which occur in about 2% of cases.6 It should be emphasized, however, that the mortality rate of this procedure in frail, elderly patients, is far from negligible, mostly due to cardiopulmonary complications. The division of the UOS muscle fibres can also be achieved by endoscopic techniques, however, by dividing the septum between the oesophageal lumen and the diverticulum using laser, cautery or stapling devices. The results listed in Table 1 show that satisfactory results can be achieved in about 90% of patients with the endoscopic approach too. This figure is very attractive as the procedure demands only about 20 minutes of general anaesthesia or, better still, it can be done under conscious sedation in the endoscopic unit. This is particularly relevant, given the intrinsic characteristics of Table 1. Results of endoscopic treatment of Zenker’s diverticula.
Author Ishioka et al (1995)12 Scher, 199618 Peracchia et al (1998)10 Baldwin and Toma (1998)19 Mulder (1999)14 Hashiba et al (1999)20 Stoecki, (2001)21 Mattinger and Ho¨rmann (2002)22 Counter et al (2002)23 a
Number of patients 42 36 92 51 167 47 30 52 31
Method
Morbidity (and mortality)
Follow-up (months)
Cautery Stapling Stapling Stapling APC laser Cautery Stapling CO2 laser Stapling
2 5a None 1 leak 12 (1)b 1 bleeding 2 mucosal tears 6 (1)c 3 leaks
38 9.3 23 15 ? ? 13.2 ? 60
Good results (%) 92.8 88.9 94.6 100 100 95.8 93.3 84.6 78.0
One perforation, one post-op fever, one transient recurrent nerve palsy, two dental injuries. Seven mediastinal emphysema, four bleeding; one death for pulmonary embolism. c Three haemorrhages, one recurrent nerve palsy, two mediastinitis; one death from myocardial infarction. b
8 M. Costantini et al
these patients. The risks of complications and even death cannot be completely overcome by these techniques, but they are lower than with open surgery, and the endoscopic treatment of Zenker’s diverticula, with its ‘minimally invasive’ nature, spread very rapidly among most surgeons, and is considered as the treatment of choice for this condition at most centres.10,11 So far there have been no prospective trials comparing the different endoscopic treatment options with surgery. Information comes from retrospective series or prospectively recorded case series, using one or more of these techniques, as in the case of all the series listed in Table 1. Moreover, given the rarity of the disease and its prevalence in elderly patients, often with severe co-morbidities, it is unlikely that randomized studies can be performed in the near future. Our recent experience consists of 24 patients treated with endoscopic stapling diverticulostomy, and 35 patients treated with surgical myotomy. Diverticulectomy was added in 21 of the latter, whereas the diverticulum was inverted below the pharyngeal muscle layer in eight patients who had a diverticulum of about 2 cm. In five patients only UOS myotomy was performed because their diverticulum was only about 1 cm. These two groups of patients have recently been reviewed.24 The choice of the operation was prompted by the size of the diverticulum and the surgical risk: patients with a high surgical risk and diverticula $ 3 cm and # 5 cm were advised to have the endoscopic procedure, whereas open surgery was recommended to younger patients at low risk, or with diverticula , 3 cm or . 5 cm. We have only sporadic experience with flexible endoscopic treatment, so the few patients concerned are not considered for the present analysis. No mortality was observed in the two groups of patients. The overall morbidity was 8.6% and all the complications were observed in the open-surgery group: there were two leakages, treated conservatively, two haemorrhages with cervical haematoma requiring drainage, and one pericarditis. The post-operative hospital stay was significantly shorter in the endoscopic group. The median duration of follow-up was 41 (range 1 –101) months. Three patients died during the follow-up for unrelated causes and two were lost to follow-up, so an adequate follow-up was obtained in 53 patients, 20 in the endoscopic group and 33 in the open surgery group. After the operation there was a statistically significant improvement in the symptom score in both groups. In the endoscopic group, however, three patients still complained of severe dysphagia and were considered as procedure failures (15%). No patients complained of severe dysphagia in the open-surgery group ðP , 0:05Þ: A posterior pouch was still evident (Figure 3) in all patients treated endoscopically and studied with post-operative barium swallow; most of the patients were asymptomatic, however. Similar results are reported by the Brussels group25 which first proposed endoscopic stapling for Zenker’s diverticula (Table 2). They retrospectively compared their experience with endoscopic treatment with both the stapler (25 patients) and the laser (54 patients), against traditional myotomy alone (seven patients), plus resection (12 patients) or pexis (36 patients), and against diverticulectomy alone (29 patients). Patients treated endoscopically had a shorter hospital stay and fasting period after the operation, as well as fewer complications (although two cervical abscesses and mediastinitis were recorded). Symptom outcome was less favourable than with open surgery, however, and only 57% of patients with a diverticulum , 3 cm were satisfied with the treatment, whereas this was true of 98% of patients treated with open surgery ðP , 0:05Þ: This difference was less evident in patients with larger diverticula (. 3 cm), however: 80 versus 96%. But still, in the series as a whole, only 75% of patients treated
Oesophageal diverticula 9
Figure 3. Barium swallow after endoscopic stapling diverticulostomy. A posterior pouch is still visible in all the patients, representing the diverticulum widely anastomosed to the oesophageal lumen.
endoscopically were symptom-free at the follow up, as opposed to 97% of the patients who had open surgery. There is only one other paper comparing endoscopic stapling diverticulostomy with open surgery in two small groups of patients (eight patients each)26; this paper emphasizes the significantly shorter operation time, post-operative hospital stay and time to oral intake, and the lower hospital charges for the patients treated
10 M. Costantini et al
Table 2. Retrospective studies comparing surgical myotomy and endoscopic treatment for Zenker’s diverticula. Surgical myotomy
Author Smith, 200226 Gutshow, 200225 Zaninotto, 200324 a b c d
Number of patients 8 67 34
Complications – 6b 5d
Endoscopic treatment
Good results (%) 100 97 100
Number of patients 8 86 24
Complications
Good results (%)
1a 3c -
100 75 87.5
p , 0:05 p , 0:05
One bleeding. Five leakages; one death for myocardial infarction. Two cervical abscesses and mediastinitis; one dental injury. Two leakages, two haematomas, one pericarditis.
endoscopically. Symptomatic relief was reportedly obtained in all patients in both groups, with only one minor complication (bleeding) in a patient of the endoscopic group. However, no details are given of symptom assessment or length of follow-up. It is difficult to draw final conclusions from these studies. Endoscopic treatment is very attractive because it is less invasive and has a lower complication rate although, when compared to surgery, it is sometimes less effective in relieving dysphagia completely. Endoscopic stapling also has other drawbacks, mainly related to the size of the diverticulum: in the case of a small diverticulum ð, 2 cmÞ the stapler anvil is too long to be properly accommodated inside the pouch and the cricopharyngeal fibres cannot be transected completely. In this sense, diverticulostomy with the laser or cautery may be more effective. On the other hand, very large diverticula ð. 5 cmÞ plunging into the mediastinum carry the risk of vascular lesions if they are transected blindly. Examples of minor drawbacks include patients who cannot open the mouth or those who cannot hyperextend the head. We observed a few such patients, where a surgical approach or the flexible endoscopic technique was used. In conclusion, Zenker’s diverticula can be effectively treated with either endoscopic diverticulostomy or open surgery. Both approaches have advantages and disadvantages. An individual approach to Zenker’s diverticula should therefore be recommended: high-risk patients with medium-size diverticula are probably better served by diverticulostomy. Open surgery should be recommended for small ð, 2 cmÞ or giant Zenker’s diverticula or in patients with a low surgical risk.
EPIPHRENIC DIVERTICULA These lesions occur less frequently than Zenker’s diverticula, with a ratio of 1:527, although this lower prevalence may be due to the fact that only a minority of these diverticula are symptomatic (about 25%). They are usually single, but it is not infrequent to find two or more synchronous diverticula (in up to 25% of cases, in our experience) as the result of the complex motor disorder behind their pathogenesis. Symptoms may vary greatly from one patient to another. Symptoms may be absent, the diverticulum
Oesophageal diverticula 11
being discovered by imaging studies performed for other reasons. When present, the most frequent symptoms are dysphagia and regurgitation of indigested food; these occur in up to 50% of the patients seeking medical attention. Retrosternal pain or heartburn, mimicking gastro-oesophageal reflux disease (GORD), may also be reported. In other instances, weight loss, halitosis, coughing and recurrent pneumonia are the presenting signs. Complications (ulcerations, bleeding) are rare these days, but spontaneous perforation (with a Boerhaave-like mechanism) may occasionally be seen. The diagnosis is radiological (Figure 4(A)): the spherical protrusion of the diverticulum from the oesophageal contour is easily identifiable. Tertiary contractions can often be seen, but the altered oesophageal motility is better documented at manometry. It should be noted, however, that standard manometry fails to document oesophageal dysmotility in about one in four patients. Prolonged 24-hour motility monitoring may help in discovering minor abnormalities.28 The treatment for these diverticula is surgical. However, the mere presence of an epiphrenic diverticulum is not, per se, an indication for surgery. The decision whether to operate must be balanced between the patient’s symptoms, the complication and operative risk, and the surgical expertise locally available, considering the rarity of the disease. A ‘masterful inactivity’, as described by Orringer29, in asymptomatic or mildly
Figure 4. (A) Barium swallow of a patient with an epiphrenic diverticulum. (B) The same patient, after laparoscopic diverticulectomy, oesophageal myotomy and Dor’s antireflux fundoplication.
12 M. Costantini et al
disturbing diverticula is good practice even if, in times of endoscopic staplers and devices, an oesophageal diverticulum may represent a tempting trophy for a hyperactive surgeon. The traditional surgical treatment for an epiphrenic diverticulum consists of an oesophageal myotomy, the diverticulectomy (or diverticulopexy), and an anti-reflux procedure, usually performed through a thoracotomy. More recently, developments in endoscopic surgical techniques have enabled the application of such techniques to the surgical treatment of diverticula of the lower oesophagus too. Thoracoscopic approach The first to be employed, as the logical consequence of the traditional thoracotomic route, was the thoracoscopic approach30 – 32, usually on the right. The patient is placed in the lateral decubitus position, and a double lumen tracheal tube is used because collapsing the lung facilitates the operation. Four ports are required: one for the camera, one for the lung retractor and two for the operating devices. With the help of intraoperative endoscopy, the diverticulum is identified and slowly isolated from the surrounding tissues with blunt dissection. Dissection is facilitated by using Babcock-like forceps applied to the diverticulum. One or two cartridges of staples are then applied to the base of the diverticulum, performing the diverticulectomy. This is a delicate step because endoscopic staplers are difficult to apply parallel to the oesophagus, at the base of the diverticulum, and the junction between two lines of staples may allow for postoperative leakage. The new articulated endostaplers may overcome this problem. An alternative approach is to perform a mini-thoracotomy, to introduce a normal linear stapler for better suture line control. The presence of the endoscope (or a 48 F bougie) inside the oesophagus prevents the lumen being narrowed excessively. The muscle layer overlying the suture line are approximated with some stitches. The oesophagus is then rotated to facilitate the execution of a myotomy on the opposite side. The proximal level of the myotomy should begin well above the level of the diverticulum neck, and continue to the oesophagogastric junction, that is identified with the help of the endoscope. Air insufflated through the scope may identify any perforation of the mucosa overlooked in performing the myotomy (hydro-pneumatic test). A fundoplication is not usually performed, because it is very difficult to perform via this approach, and may be not considered strictly necessary, given the minimal dissection performed at the oesophago-gastric junction, and the potent anti-reflux drugs available today. Laparoscopic approach More recently, increasing confidence with the laparoscopic approach to the hiatal region for the treatment of GORD and achalasia, and the accessibility of the lower thoracic oesophagus through the hiatus, have prompted the use of the laparoscopic approach for the surgical treatment of epiphrenic diverticula.33 The position of the patient and surgeon, and the trocars sites are the same as for the laparoscopic treatment of functional diseases of the esophagogastric junction. The phreno – oesophageal membrane is divided, and the diaphragmatic crura exposed. The oesophagus is then isolated and encircled with a silicon tube for safe traction. Blunt dissection of the lower mediastinum is then begun, paying attention to remain on the oesophageal wall, until the diverticular pouch is identified and carefully isolated up to the superior margin of its neck. Peri-operative endoscopy is very helpful in this step. When the diverticulum is completely isolated, the diverticular neck is sutured and resected with a linear endoscopic stapler, with the endoscope or a bougie inside the oesophageal lumen to
Oesophageal diverticula 13
Figure 5. Laparoscopic diverticulectomy: after complete mobilization of the diverticulum, diverticulectomy is performed by firing one or two endostapler cartridges at the base of the diverticulum. The endoscope or a bougie are inserted into the oesophagus, to avoid narrowing of the lumen. With this approach, the instrument is parallel to the oesophageal axis. After the diverticulectomy, the procedure is completed by a myotomy (on the opposite side of the oesophagus) and a partial antireflux procedure.
avoid narrowing the lumen (Figure 5). The resected diverticulum is then removed through a port, and the sutured line is protected by over sewing the muscular edges with some interrupted or continued suture. A hydropneumatic test can be performed by inflating air through the scope in the oesophageal lumen to identify any leakages. The myotomy is then performed on the opposite side of the oesophagus, from a level well above the upper margin of the diverticulum and down to 1 –2 cm below the cardia, on the gastric side. The operation is then completed with a partial anterior fundoplication (following Dor’s technique). These last two steps are identical to those of laparoscopic treatment for oesophageal achalasia. Results and discussion Epiphrenic diverticula are rare. They are often asymptomatic and do not always require surgical treatment. A group of asymptomatic patients was followed up at the Mayo Clinic for a median of 9 years and none became symptomatic during said time.27 Other authors28 favour the treatment of such patients, based on the finding of sometimes lifethreatening pulmonary complications in 25– 45% of the cases they followed. However, the success rate with the traditional thoracotomic approach is only about 85% and the operation carries a considerable morbidity (namely leakages from the suture line) and, above all, an undeniable mortality. The basic principles of surgical treatment for
14 M. Costantini et al
epiphrenic diverticula were defined back in the 1960s and have changed little since. Areas of controversy still exist34, but most surgeons agree that complete relief of any distal oesophageal obstruction (with myotomy) is the mainstay of the treatment, even if a motor disorder cannot be found with function studies. The introduction of minimally invasive surgery certainly revitalized interest in this pathology, as in the case of several other surgical fields. A couple of successful case reports of thoracoscopic resection of epiphrenic diverticula, with31 or without32 myotomy, have recently been reported. Another study, using thoracoscopic diverticulectomy, alone or associated with endoscopic pneumatic dilation of the cardia in the case of evident motor abnormalities, reported a high incidence of post-operative leakages, prompting the authors to abandon such an approach.30 We have personally treated only one patient with this approach, performing a simple thoracoscopic diverticulectomy after forceful pneumatic dilation of the cardia. In this particular patient, we observed a radiological recurrence of the diverticulum, albeit asymptomatic, a year later. These experiences confirm the importance of oesophageal myotomy in treating such complex disorders and the need to avoid modifying a consolidated surgical technique simply to render it more suitable for a novel approach. A completely laparoscopic approach was proposed by Rosati et al in 199833, and in 2001 the same group reported on their whole experience in 11 patients treated in this way, with a median follow-up of 3 years.35 Their results were good in all the patients treated, with only one post-operative leakage that required a thoracotomy to be repaired. Other reports with this technique have been published, reporting the successful laparoscopic treatment of epiphrenic diverticula36,37, even after an iatrogenic perforation.38 Our personal experience consists of eight cases (Table 3). Diverticulectomy (with one or two endostapler cartridges), extramucosal myotomy on the opposite side of the oesophagus and anterior fundoplication were completed laparoscopically in all the patients. Morbidity was high, however, with three postoperative leakages, requiring prolonged naso-gastric suction, parenteral nutrition and antibiotics. All three patients healed with this conservative treatment. After a median follow-up of 36.5 months, all the eight patients are satisfied with the treatment and are eating normally, but persistent reflux symptoms make continuous Proton Pump Inhibitors (PPI) treatment necessary in two of them. For the time being, the relatively recent introduction of such techniques—and the rarity of the disease—do not provide sufficient experience to enable us to draw definitive conclusions, such as those which can now be drawn, for example, for the laparoscopic Table 3. Personal experience with the laparoscopic treatment of epiphrenic diverticula. Patient number
Sex
Age
Size
Manometry
Morbidity
Follow-up (months)
Result
1 2 3 4 5 6 7 8
M F F F M M M F
64 70 59 41 54 55 67 41
4 cm 4 cm 5 cm 4 cm 7 cm 6 cm 5 cm 7 cm
LOS and body dysmotility LOS and body dysmotility Body dysmotility Normal Body dysmotility Body dysmotility Achalasia Normal
– – Pleural effusion – Leakage Leakage – Leakage
51 45 40 40 33 29 24 18
Good Good Good Good GORD Good Good GORD
Oesophageal diverticula 15
treatment of GORD or achalasia. Some considerations are nonetheless worth stating. A laparoscopic approach to these diverticula is certainly feasible. It offers a good access to the mediastinal oesophagus and therefore makes all three steps of the consolidated surgical procedure (myotomy, diverticulectomy and fundoplication) feasible in a single session. It may therefore become the modern treatment of choice for these diverticula. Among its advantages, it is worth emphasizing the fact that the endoscopic stapler lies parallel to the axis of the oesophagus (and therefore also to the neck of the diverticulum). Further, the optimal visualization of the hiatal region facilitates myotomy on both the oesophageal and gastric sides, and makes anterior fundoplication easy. Further experience is needed with larger numbers of patients to confirm the validity of this approach, but the rarity of the disease, the relative difficulty of the treatment and its considerable, potentially severe morbidity make this treatment suitable only for specialized referral centres.
Practice points † Zenker’s and epiphrenic diverticula of the oesophagus are caused by a disordered oesophageal motility † surgical treatment of oesophageal diverticula requires the elimination of functional obstruction by means of myotomy † myotomy of the UOS can effectively be achieved by means of endoscopic techniques † minimally invasive surgery has been successfully employed for the treatment of oesophageal epiphrenic diverticula † in this context, laparoscopic access seems more promising than thoracoscopic access † because of the rarity of the disease, laparoscopic treatment of epiphrenic diverticula must be reserved for dedicated centres with expertise in laparoscopic surgery of the gastro-oesophageal junction and a good knowledge of oesophageal pathophysiology
Research agenda † further studies comparing endoscopic treatment of Zenker’s diverticula to open surgery are needed in order to clarify the indications of this technique † the laparoscopic treatment of epiphrenic oesophageal diverticula requires further evaluation in order to verify its efficacy and safety. Until more experience is acquired, this approach should not be considered the gold standard of therapy for such diverticula
REFERENCES 1. Posthletwait RW. Diverticula of the esophagus. Surgery of the Esophagus. Norwalk: Appleton Century Crofts, 1986, pp 129 –160. 2. Killian G. The mouth of the esophagus. Laryngoscope 1907; 17: 421– 428. 3. Ludlow A. A case of obstructed deglutition from a preternatural dilatation of, and a bag formed, in the pharynx. Medical Observations and Inquiries of the Society of Physicians 1760; 3: 85– 101.
16 M. Costantini et al 4. Zenker FA & von Ziemssen H. Krankheiten des Oesophagus. In von Ziemsen H (ed.) Handbuch der Speciellen Pathologie und Therapie, 7. FC Vogel: Leipzig, 1877. * 5. Cook IJ, Gabb M & Panagopoulos V. Zenker’s diverticulum is a disorder of upper esophageal sphincter opening. Gastroenterology 1993; 103: 1229–1235. 6. Payne WS. The treatment of pahryngoesophageal diverticulum: the simple and the complex. Hepatogastroenterology 1992; 39: 109 –114. 7. Dohlman G & Mattson O. The endoscopic operation for hypopharyngeal diverticula. Archives of Otolaryngology 1960; 71: 744–752. 8. Knegt PP, de Jong PC & van der Schans EJ. Endoscopic treatment of the hypopharyngeal diverticulum with CO2 laser. Endoscopy 1985; 17: 205–206. * 9. Collard JM, Otte IB & Kestens PJ. Endoscopic stapling technique of esophagodiverticulostomy for Zenker’s diverticulum. Annals of Thoracic Surgery 1993; 56: 573–576. * 10. Peracchia A, Bonavina L, Narne S et al. Minimally invasive surgery for Zenker’s diverticulum. Analysis of results in 95 consecutive patients. Archives of Surgery 1998; 133: 695–700. 11. Omote K, Feussner H, Stein HJ et al. Endoscopic stapling diverticulostomy for Zenker’s diverticulum. Surgical Endoscopy 1999; 13: 535–538. 12. Ishioka S, Sakai P, Maluf FF & Melo JM. Endoscopic incision of Zenker’s diverticula. Endoscopy 1995; 27: 433–437. 13. Mulder CJJ, Robijn RJ & Thies JE. Flexible endoscopic treatment of Zenker’s diverticulum: a new approach. Endoscopy 1995; 27: 438–442. 14. Mulder CJJ. Zapping Zenker’s diverticulum: gastroscopic treatment. Canadian Journal of Gastroenterology 1999; 13: 405–407. 15. van Overbeek J, Hoeksema PE & Edens ET. Microendoscopic surgery of the hypopharyngeal diverticulum using electrocoagulation or carbon dioxide laser. Annals of Otology, Rhinology and Laryngology 1984; 93: 34–36. * 16. Mulder CJJ, Costamagna G & Sakai P. Zenker’s diverticulum: treatment using a flexible endoscope. Endoscopy 2001; 33: 991–997. * 17. Sideris L, Qi Chen L, Ferraro P & Duranceau AC. The treatment of Zenker’s diverticula: a review. Seminars in Thoracic and Cardiovascular Surgery 1999; 11: 337 –351. 18. Scher RI & Richtsmeier WJ. Long-term experience with endoscopic stapled-assisted esophagodiverticulostomy for Zenker’s diverticulum. Laryngoscope 1998; 108: 200– 205. 19. Baldwin DL & Toma AG. Endoscopic stapled divertciulostomy: a real advance in the treatment of hypopharyngeal diverticulum. Clinical Otolaryngology 1998; 23: 244–247. 20. Hashiba K, de Paula AL, da Silva JG et al. Endoscopic treatment of Zenker’s diverticulum. Gastrointestinal Endoscopy 1999; 49: 93–97. 21. Stoecki SJ & Schmid S. Endoscopic stapled-assisted diverticulo-esophagostomy for Zenker’s diverticulum: patient satisfaction and subjective relief of symptoms. Surgery 2002; 131: 158–162. 22. Mattinger C & Ho¨rmann K. Endoscopic diverticulostomy of Zenker’s diverticulum: management and complications. Dysphagia 2002; 17: 34 –39. 23. Counter PR, Hilton ML & Baldwin DL. Long-term follow-up of endoscopic stapled diverticulotomy. Annals of the Royal College of Surgeons of England 2002; 84: 89–92. * 24. Zaninotto G, Narne S, Costantini M et al. Tailored approach to Zenker’s diverticula. Surgical Endoscopy 2003; 17: 129 –133. * 25. Gutschow CA, Hamoir M, Rombaux PH et al. Management of pharyngoesophageal (Zenker’s) diverticulum: which technique? Annals of Thoracic Surgery 2002; 74: 1677– 1683. 26. Smith SR, Genden EM & Urken ML. Endoscopic stapling technique for the treatment of Zenker diverticulum vs standard open-neck technique. Archives of Otolaryngology–Head and Neck Surgery 2002; 128: 141 –144. * 27. Benacci JC, Deschamps C, Trastek VF et al. Epiphrenic diverticulum: results of surgical treatment. Annals of Thoracic Surgery 1993; 55: 1109–1113. * 28. Nehra D, Lord RV, DeMeester TR et al. Physiologic basis for the treatment of epiphrenic diverticulum. Annals of Surgery 2002; 235: 346 –354. 29. Orringer MB. Epiphrenic diverticula: fact and fable. Annals of Thoracic Surgery 1993; 55: 1067– 1068. 30. Peracchia A, Bonavina L, Rosati R & Bona S. Thoracoscopic resection of epiphrenic esophageal diverticela. In Peters JH & DeMeester TR (eds) Minimally Invasive Surgery of the Foregut. St Louis: Quality Medical Publishing, 1994, pp 110– 116. 31. Saw EC, McDonald TP & Kam NT. Video-assisted thoracoscopic resection of an epiphrenic diverticulm with esophagomyotomy and partial fundoplication. Surgical Laparoscopy and Endoscopy 1998; 8: 145–148. 32. Stuart RC, Wyman A, Chan AW et al. Thoracoscopic resection of oesophageal diverticulum: a case report. Journal of the Royal College of Surgeons of Edinburgh 1996; 41: 118–119.
Oesophageal diverticula 17 * 33. Rosati R, Fumagalli U, Bona S et al. Diverticulectomy, myotomy and fundoplication through laparoscopy: a new option to treat epiphrenic esophageal diverticula? Annals of Surgery 1998; 227: 174–178. 34. Lerut T & Hiebert CA. Esophageal diverticula. In Pearson FG (ed.) Esophageal Surgery. New York: Churchill Livingstone, 2002, pp 507– 514. 35. Rosati R, Fumagalli U, Bona S et al. Laparoscopic treatment of epiphrenic diverticula. Journal of Laparoendoscopic and Advanced Surgical Techniques 2001; 11: 371–375. 36. Chami Z, Fabre JM, Navarro F & Domergue J. Abdominal laparoscopic approach for thoracic epiphrenic diverticulum. Surgical Endoscopy 1999; 13: 164–165. 37. Eubanks TR & Pellegrini CA. Minimally invasive treatment of esophageal diverticula. Seminars in Thoracic and Cardiovascular Surgery 1999; 11: 363 –367. 38. Tinoco RC, Tinoco AC & El-Kadre L. Perforated epiphrenic diverticulum treated by video laparoscopy. Surgical Endoscopy 1999; 13: 270–272.