Fluoroscopically-guided palliative stenting for the management of malignant oesophageal obstruction: A five year experience

Abstract

Introduction

Oesophageal cancer is the eighth most common cancer in the world and in 2008 there were an estimated 482,000 new cases with a mortality rate is greater than 80%.1,2 Early diagnosis is difficult as symptoms of dyspepsia and reflux disease are highly prevalent and non-specific. Sinister presenting symptoms associated with malignancy include chronic gastrointestinal bleeding, unintentional weight loss, iron deficiency anaemia, dysphagia, vomiting and an abdominal mass.2

Dysphagia is often a late presenting symptom and usually indicates more advanced stages of malignancy.3 More than 50% of patients have non-operable disease at time of presentation and 90% of these patients will require a procedure to relieve dysphagia.2 Dysphagia is both nutritionally and psychologically debilitating and its onset heralds a precipitous decline in the quality of life of patients with oesophageal cancer. Available modalities to treat malignant oesophageal obstruction include self-expanding metal stents (SEMS), argon plasma anticoagulation (APC) and brachytherapy.

Multiple studies have shown that SEMS have good outcomes, with low morbidity and re-intervention rates.4–11 In the UK, SEMS are the most commonly employed method of palliating oesophageal obstruction. This study aims to evaluate the outcomes of patients treated for malignant oesophageal obstruction with SEMS over a five-year period, including outcomes, procedure-related complications and survival.

Methods

All patients undergoing insertion of SEMS from 1st January 2006 to 31st December 2010 were identified from a prospectively maintained interventional radiology database at Frimley Park Hospital (Frimley, UK). Stents performed for benign disease were excluded. Data were obtained retrospectively from patient case notes and electronic records.

The data retrieved included patient demographics, tumour histology and stage, procedural details, early and late complications, length of stay, re-interventions/readmissions, length of survival and cause of mortality. Information regarding palliative radiotherapy and chemotherapy was also collected.

In all cases the indication for SEMS was the development of dysphagia to solids or progression to complete dysphagia. All patients had non-operable disease, defined due to disease-specific factors, such as advanced local staging or metastatic disease, or because patients were not fit to undergo major resection due to co-morbidities.

All procedures were performed under conscious sedation by one of two experienced consultant interventional gastrointestinal radiologists in a dedicated radiology intervention suite, using a combination of midazolam and fentanyl. Iodinated oral contrast examination was performed to assess the margins of the tumour and degree of stenosis. A soft guidewire was inserted under fluoroscopic guidance past the stenosis. Either single or overlapping 12 or 15 cm, 18 mm diameter double layer stents were deployed, using an over-the-wire delivery system (Taewoong Niti-S; Taewoong Medical, Gimpo, Korea). The inner polyurethane covered layer can prevent ingrowth of tumour, which occurs in uncovered stents and the uncovered outer nitinol wire layer prevents displacement, which occurs in covered stents.5

Following the procedure, dietary intake was re-established under the supervision of a dietician, including provision of verbal and written instructions. Clinical success of the procedure was defined by a resolution of dysphagia and patients tolerating soft diet. After discharge they were followed up by either a telephone consultation by a specialist nurse or outpatient review after approximately six weeks.

Non-parametric data are presented as median (interquartile range, IQR). Survival from the date of diagnosis and date of stent insertion was estimated using the Kaplan-Meier method. The analysis was performed using the IBM SPSS Statistics 20.0 software (IBM Co., Armonk, NY, USA).

Results

A total of 120 SEMS procedures were performed on 109 patients. There were 66 men and 43 women giving a male vs female ratio of 1:1.65 and the median age was 76 years (IQR, 69–82 years). Tumour histology was available for 90 patients and 24 patients (22%) had a squamous cell histopathology, 65 patients (60%) had adenocarcinoma and one patient had an anaplastic tumour.

Staging details were available for 105 patients and 48 patients (46%) had stage T3 and 57 patients (54%) had T4 tumour grading.12 All patients had lymph node involvement and 48 patients (46%) had metastatic disease. A single patient had recurrent disease at the level of the anastomosis site. The patient cohort was not fit for curative treatment because of extensive disease or poor fitness for surgery. Seventy-eight of these patients (74%) had palliative chemotherapy prior to stent insertion.

SEMS was performed at a median of 4.4 months (IQR, 0.9–9.8 months) from initial diagnosis. A satisfactory stent position was obtained in all patients. Although all patients had immediate stent patency confirmed by passage of oral contrast, 4 patients (3%) did have delays in establishing oral intake following the procedure. Three patients required two overlapping stents; the other 106 patients required either a 12 or 15 cm stent. All patients were successfully discharged with a median length of stay of 3 days (IQR, 2–7 days). There were no immediate complications or perforations at time of procedure. Ten patients (9%) required re-stenting for recurrent dysphagia, including 3 patients (3%) within 30 days of the initial procedure.

Thirty-two patients (29%) developed early complications and 20 patients (18%) died within 30 days (Fig. 1). Of the patients that died, 18 patients (17%) had a gradual deterioration. These patients all exhibited signs of respiratory and/or renal failure, before succumbing. Two patients (2%) had massive gastrointestinal bleeds directly from their tumours and died despite endoscopic intervention. Four patients had delayed oral intake, requiring re-stenting (3 patients) or an endoscopy to relieve food bolus obstruction (1 patient). Respiratory complications included four patients with pneumonia and a patient who developed a tracheo-bronchial fistula. In total, there were 12 patients (11%) with stent-related complications. Of 12 patients developing dysphagia due to tumour ingrowth and stent obstruction, 10 patients (9%) were re-stented and 2 patients (2%) were managed conservatively.

Overall median survival from the date of diagnosis was 9.4 months (IQR, 4–15.4 months) with an actuarial survival of 37% at 1 year and 9% at 2 years. The median survival from SEMS procedure was 2.7 months (IQR, 1.5–5.6 months) with an actuarial survival of 21% at 6 months and 7% at 1 year. These findings are displayed in Fig. 2. Only 11 patients underwent chemotherapy post-stenting, with a median survival of 3.7 months. One patient underwent post-stenting radiotherapy and survived for 5.5 months.

Discussion

We have performed a large number of SEMS procedures at our district general hospital with excellent technical success rates. There were no serious direct procedure-related complications resulting in death, which has been reported in previously published case series. There have been two studies reporting death of patients from significant gastrointestinal haemorrhage.6,7,9–11,13,14 Our data compares well in number of parameters with the published data. Eleven percent of patients suffered stent-related complications, including re-stenosis and bleeding due to vessel erosion, compared with published rates of 10% to 70%.6,7,9–11,13,14 Overall complication rates of 29% are less than the reported rates of 38% to 43% elsewhere.7–11,15 Our study has a stent re-insertion rate of 9%, which is at the lower end of the reported range of 7% to 26%.6,9,13,14,16,17 Post-stent median survival is dismal, but unfortunately within the expected range for patients undergoing palliative treatment for oesophageal cancer. We believe this report demonstrates that good outcomes can be achieved with fluoroscopic-guided stent insertion by experienced radiologists. Post-stent care was guided by a multi-disciplinary team, which included dieticians.

There have been key review articles and a Cochrane meta-analysis discussing the best treatment options for oesophageal obstruction.13,17,18 SEMS are established as superior treatments to plastic tubes. Further study is required to show whether covered or uncovered stents are preferable.17,19 Our unit policy has been to use double layer stents, with an uncovered outer layer, as these have low displacement rates (0% in this series) and good technical outcomes, in terms of stent position and immediate post-procedure patency.5 Thermal and ablative techniques, such as APC, have also been used but are not as widely available. APC gives good initial results by reducing tumour bulk, but re-intervention rates are high, especially due to bleeding complications. Brachytherapy is another effective palliation and sustained dysphagia relief is comparable to SEMS.18,20,21 However, availability of brachytherapy is very limited in the UK and USA.22 In future, combination treatments with stents and brachytherapy may become the mainstay of palliation and promising trials are underway.20,23,24

Oesophageal cancer is a debilitating illness with reports as high as 80% of patients presenting with weight loss.25,26 This relates to actual dysphagia-related anorexia and cancer cachexia.14 Cachexia is characterised by involuntary weight loss associated with reduced muscle mass, generalised weakness, oedema and deteriorating physical and mental function.14 Cancer cachexia syndrome does not respond to caloric supplementation and heralds a poor prognosis.14,27,28 Lecleire et al14 reviewed 120 patients undergoing SEMS procedures for oesophageal malignancy and found that, even after successful intervention, there was still a significant decrease in body mass index, albumin levels and World Health Organisation performance status scores. They concluded that this was due to the underlying disease process and, despite the ability to swallow food, patients were still markedly malnourished. These three parameters were independent predictors of 30-day mortality. These findings have been corroborated in further published studies.8,14,29,30

Our study demonstrates that SEMS effectively relieves dysphagia. As a one-off palliative procedure, it is important to have low levels of peri-procedural morbidity and minimal re-interventions to maximise quality of life and minimise re-hospitalisation.

Five patients had respiratory complications post-stenting. Peri-procedural aspiration pneumonia may be compounded by pre-existing smoking-related lung disease, which is common in patients with oesophageal cancer. The combination of respiratory complications with cancer cachexia often advances the onset of the terminal phase of illness. Eighteen patients died within 30 days of the procedure of a global gradual deterioration and some or all of these patients will have had a degree of respiratory compromise. Median post-stent survival was only 2.7 months, highlighting the importance of involving palliative care services as soon as practicable. We have used the results of this study to introduce robust referral protocols in place to initiate fast-track assessment by the hospital palliative care team.

There are limitations to this study. Data collection was performed retrospectively and was limited to palliation by one treatment modality of SEMS alone. Patients with dysphagia who were too unwell to have a stent were not included in this study. No comparison groups of patients undergoing other palliative procedures, such as APC or brachytherapy are available, as these treatments are not offered within our regional cancer network. We did not employ an objective scoring system for assessment of dysphagia, which has been utilised in previous studies.6,10,18,20 However, the hospital notes were clearly indicative of whether SEMS was successful or not in relieving dysphagia. Further studies should include quality of life assessment tools, which may give rise measures of patient-assessed outcomes. These patient-centred outcomes will help in the design of future comparative studies of SEMS with other treatments. This study demonstrates that SEMS is an effective and safe method of palliation for patients with advanced or non-operable oesophageal cancer. Survival post-stent-ing is short and palliative treatments and patient support should be initiated as soon as the diagnosis has been made.

Article information

References