European multicentre experience of staged hybrid atrial fibrillation ablation for the treatment of persistent and longstanding persistent atrial fibrillation.

The management of non-paroxysmal atrial fibrillation (AF) remains controversial. We examined the efficacy and safety of the 2 stage Hybrid AF ablation approach by analysing the largest series of this technique reported so far. METHODS: The approach aims to electrically isolate the left atrial posterior wall incorporating the pulmonary veins ('box-set'pattern). An initial video-assisted thoracoscopic (VATS) epicardial ablation is followed after a minimum of 8 weeks by endocardial radiofrequency catheter ablation. RESULTS: Of 175 patients from 4 European cardiothoracic centers, who underwent the surgical (COBRA Fusion, AtriCure Inc) 1st stage ablation, 166 went on to complete 2nd stage catheter ablation. At median follow up of 18 months post 2nd stage procedure 93/166 (56%) had remained free of AF or atrial tachycardia (AT) recurrence off antiarrhythmic drugs. 110/175 62.9% were in sinus rhythm off all antiarrhythmic drugs at last clinic follow-up (132/175 75.4% including those on antiarrhythmic drugs). 18 patients (10.8%) underwent a further re-do ablation (mean of 1.1 ablations per patient) 105/166 (63%) remained free of AF/AT recurrence off antiarrhythmic drugs following last ablation procedure.Latterly, ILRs have been implanted in patients (n = 56); 60% have remained fully arrhythmia free and 80% have shown AF burden < 5% at a median 14 months follow-up [IQR: 13.5 (8-21.5)]. Only 10.9% have reverted to persistent AF. 5 patients (2.9%) had a perioperative stroke and 4 patients (2.3%) exhibited persistent weakness of the right hemidiaphragm following stage 1 VATS epicardial ablation. One patient died following stroke (overall mortality 0.6%). CONCLUSIONS: In patients with non-paroxysmal AF with unfavourable characteristics for catheter ablation, the staged hybrid approach results in acceptable levels of freedom from recurrent atrial arrhythmia, however, complication rates are higher than with catheter ablation alone.

The Late Electrophysiological Consequences Of Posterior Wall Isolation In Patients With Atrial Fibrillation.

INTRODUCTION: There are many different lesion sets that are used for the surgical ablation of atrial fibrillation (AF). One such pattern is the 'box set', a single ring of scar delivered anterior to the pulmonary veins, which aims to electrically isolate the posterior wall from the rest of the heart. However it remains unclear whether posterior wall isolation (PWI) is an effective lesion set for maintenance of sinus rhythm and whether it is necessary to achieve complete bidirectional block. We investigated the long-term integrity of the 'box set' lesion created during surgical AF ablation by epicardial High Intensity Focussed Ultrasound (HIFU). All patients had documented persistent or recurrent paroxysmal AF prior to surgery. We correlated this with subsequent success or failure in the abolition of atrial fibrillation. METHODS: With regional ethical and R&D approval, 101 patients who had previously undergone HIFU AF ablation greater than 4 years ago were screened for inclusion in the study. 17 patients agreed to late electrophysiological study: 11 with on-going AF and 6 in normal sinus rhythm. Clinical history and 7-day holters were used to define the NSR group. We performed a diagnostic EP study using a transseptal approach in fully anticoagulated patients (INR>2.0 and ACT maintained at >300s). A catheter was placed in the coronary sinus (CS) and a circular multipolar mapping catheter was used to map the left atrium and pulmonary veins. Patients in atrial fibrillation were cardioverted. We recorded whether posterior wall (PW) and pulmonary vein (PV) isolation had been achieved at the surgical procedure. In selected cases we recorded a voltage map using either CARTO (Biosense- Webster) or NavX (St Jude Medical) to identify areas of ablation scar. RESULTS: All 11 patients with AF had absence of PW+PV isolation with fractionated electrograms recorded across the PW. In the 6 patients with long-term freedom from AF, PW+PV isolation was confirmed in 4 (67%) and in 1 there was prolonged conduction across the box-set lesion with CS to PW activation time of around 200ms versus 45ms from mid-CS to left atrial appendage. Of the 4 patients with confirmed PW+PV isolation, 1 had dissociated spontaneous atrial potentials within the box set area and the other 3 had electrical silence throughout with inability to capture the posterior wall pacing at 10mA at multiple sites. CONCLUSIONS: There appears to be a clear correlation between the successful restoration of long-term sinus rhythm and isolation / delayed conduction from the pulmonary veins and posterior wall. Given the advent of hybrid atrial fibrillation ablation techniques designed to deliver this lesion set, these findings are timely and highly relevant.

What is the impact of endoscopic vein harvesting on clinical outcomes following coronary artery bypass graft surgery?

Objective Endoscopic vein harvesting (EVH) is increasingly used as an alternative to open vein harvesting (OVH) for coronary artery bypass graft (CABG) surgery. Concerns about the safety of EVH with regard to midterm clinical outcomes following CABG have been raised. The objective of this study was to assess the impact of EVH on short-term and midterm clinical outcomes following CABG. Design This was a retrospective analysis of prospectively collected multi-centre data. A propensity score was developed for EVH and used to match patients who underwent EVH to those who underwent OVH. Setting Blackpool Victoria Hospital, Plymouth Derriford Hospital and the University Hospital of South Manchester were the main study settings. Patients There were 4709 consecutive patients who underwent isolated CABG using EVH or OVH between January 2008 and July 2010. Main outcome measures The main outcome measure was a combined end point of death, repeat revascularisation or myocardial infarction. Secondary outcome measures included in-hospital morbidity, in-hospital mortality and midterm mortality. Results Compared to OVH, EVH was not associated with an increased risk of the main outcome measure at a median follow-up of 22 months (HR 1.15; 95% CI 0.76 to 1.74). EVH was also not associated with an increased risk of in-hospital morbidity, in-hospital mortality (0.9% vs 1.1%, p=0.71) or midterm mortality (HR 1.04; 95% CI 0.65 to 1.66). Conclusions This multi-centre study demonstrates that at a median follow-up of 22 months, EVH was not associated with adverse short-term or midterm clinical outcomes. However, before the safety of EVH can be clearly determined, further analyses of long-term clinical outcomes are required.

Pulmonary metastatic melanoma -- the survival benefit associated with positron emission tomography scanning.

OBJECTIVE: Positron emission tomography (PET) scanning is more sensitive at detecting metastatic disease than conventional radiological techniques. For patients with pulmonary metastatic melanoma, we investigate if PET scanning to detect occult extra pulmonary disease prior to thoracotomy and metastectomy is associated with improved survival compared to patients staged by conventional radiology. METHODS: Between November 1984 and December 1999, 121 patients (90 males, 31 females) have undergone a thoracotomy and pulmonary metastectomy for metastatic melanoma. The age range was 19-84 years (mean 57, median 59). In every case all palpable nodules were removed and the diagnosis confirmed histologically. A total of 68 (56%) patients had a PET scan preoperatively, 53 (44%) underwent conventional or nuclear imaging. Patients with only radiologically isolated pulmonary disease are included. RESULTS: Survival is 100% complete and totals 238 pt/years (mean 2.2 years, median 1.4 years). Survival (+/-SE) at 1, 3, 5 and 7 years for all patients is 68% (+/-4.5) (n=67), 36.6% (+/-5.2) (n=27), 22.1% (+/-4.8) (n=15) and 13.5% (+/-4.2) (n=7), respectively. Survival (+/-SE) was significantly better at 3 and 5 years in patients who underwent a PET scan preoperatively (Log rank P=0.002). There was no significant difference in survival by 7 years. CONCLUSIONS: There is a significant survival benefit associated with excluding extra pulmonary disease using a PET scan prior to thoracotomy and metastectomy. We recommend that PET scanning be used in the investigation of patients with pulmonary metastatic melanoma prior to metastectomy.

Autotransfusion decreases blood usage following cardiac surgery -- a prospective randomized trial.

INTRODUCTION: 10% of blood issued by the National Blood Service (220,000) is utilised in cardiac procedures. Transfusion reactions, infection risk and cost should stimulate us to decrease this transfusion rate. We tested the efficacy of autotransfusion of washed postoperative mediastinal fluid in a prospective randomized trial. PATIENTS AND METHODS: 166 patients undergoing coronary artery bypass grafting (CABG), valve or CABG + valve procedures were randomized into three groups. The indication for transfusion was a postoperative haemoglobin (Hb) < 10 g/l or a packed cell volume (PCV) < 30. When applicable, group A patients received washed post-operative drainage fluid. Group B all received blood processed from the cardiopulmonary bypass (CPB) circuit following separation from CPB and if appropriate washed post-operative drainage fluid. Group C were controls. Groups were compared using analysis of variance. RESULTS: There was no significant difference in age, sex, type of operation, CPB time and preoperative Hb and PCV between the groups. Blood requirements were as shown. [table - see text] Twelve patients in group A and 10 in group B did not require a homologous transfusion following processing of the mediastinal drainage fluid. CONCLUSION: Autotransfusion of washed postoperative mediastinal fluid can decrease the amount of homologous blood transfused following cardiac surgery. There was no demonstrable benefit in processing blood from the CPB circuit as well as mediastinal drainage fluid.

Open commissurotomy for critical isolated aortic stenosis in neonates.

BACKGROUND: The optimal management of critical aortic stenosis in early infancy remains controversial. The aim of this study was to assess the early and late outcomes following open surgical valvotomy for critical aortic stenosis in neonates and to provide a framework of data against which current results of other treatment approaches can be evaluated. METHODS: Eighteen consecutive neonates (mean age 9.2 days, range 1 to 26 days) undergoing an open valvotomy for critical isolated aortic stenosis (the standard treatment for this condition in our unit) between 1984 and 2000 were studied. The mean aortic valve gradient was 79.4 mm Hg. Twelve neonates received prostaglandins and 10 received inotropic agents preoperatively. Follow-up was complete (mean 8.1 years, range 1 month to 15 years). RESULTS: There was no operative mortality. At discharge, the mean aortic valve gradient was 37.2 mm Hg, with 6 patients having mild and 2 having moderate aortic regurgitation. Six patients required a reoperation; 3 of these had an aortic valve replacement at 9 to 11 years of age. Kaplan-Meier 5- and 10-year freedoms from any aortic reoperation or reintervention were 85 and 55%, respectively; 5- and 10-year freedoms from aortic valve replacement were 100 and 79%, respectively. A 14-year-old boy died from endocarditis 4 years following an aortic valve replacement in another unit. Kaplan-Meier 10-year survival was 100%. All survivors are in New York Heart Association I class and are leading normal lives. Their mean aortic valve gradient is 34.5 mm Hg, and none has significant aortic regurgitation. CONCLUSIONS: Open valvotomy for critical aortic stenosis in neonates carries a low operative risk and provides lengthy freedom from recurrent stenosis or regurgitation. Reoperations are inevitable, but aortic valve replacement can be delayed until the implantation of an adult-sized prosthesis is possible. Late survival is excellent. We consider open surgical valvotomy to be the treatment of choice for critical neonatal aortic stenosis.

Endoscopic vein harvesting with the aid of carbon dioxide insufflation.

Endoscopic harvesting of the long saphenous vein has been introduced to decrease the morbidity of obtaining venous conduit for coronary artery bypass grafting. Herein is described an endoscopic method using carbon dioxide insufflation into the tissues around the vein. This has several advantages; improved vision, no physical retraction required, easier development of tissue planes, and improved hemostasis.

Aortic valve replacement for endocarditis: determinants of early and late outcome.

BACKGROUND AND AIM OF THE STUDY: The study aim was to determine risk factors for operative mortality, recurrent infection, reoperation and long-term survival following aortic valve replacement (AVR) for infective endocarditis. METHODS: Between 1973 and 1997, 109 patients (91 male, 18 female, mean age 52.6 years) underwent isolated AVR for infective endocarditis in our unit. Native valve endocarditis was present in 89 (81.6%) and prosthetic valve endocarditis in 20 (18.4%). Active culture-positive endocarditis was present in 53 (48.6%). Preoperatively, 99 patients (90.8%) were in NYHA classes III and IV. Indications for surgery included cardiac failure in 41 patients, valvular dysfunction in 38, vegetations in 18, sepsis in seven, abscess in six and embolism in four. Mechanical valves were implanted in 69 patients (63.3%) and bioprostheses in 40 (36.7%), including a homograft in 19 (17.4%). Follow up was complete (mean 5.8 years; range: 0-23.8 years; total 633.5 patient-years). RESULTS: The operative mortality was 10.1% (11 deaths). At ten years, freedom from recurrent infection was 94.2%, and freedom from reoperation 83.6%. Biological valve and younger age were significant adverse parameters for freedom from reoperation (p = 0.01 and p = 0.01). There have been 21 late deaths, 15 due to cardiac causes. Kaplan-Meier survival, including operative mortality, at five and ten years was 77.4% and 68.0%, respectively. On Cox proportional hazards regression, Staphylococcus aureus infection (p = 0.008) and older age (p = 0.04) were independent adverse predictors of survival. CONCLUSION: AVR for endocarditis carries a relatively high operative mortality, but can result in a satisfactory freedom from recurrent infection, reoperation and long-term survival. Analysis of our series demonstrates that implantation of a biological valve limits the freedom from reoperation and that infection by Staph. aureus reduces the probability of long-term survival.

Ventricular thrombosis and systemic embolism in bodybuilders: etiology and management.

Increased thrombogenicity and acute embolism are well-recognized complications of chronic anabolic steroid abuse. The following cases highlight such dangers in steroid-enhanced bodybuilders who developed intracardiac thrombosis that subsequently embolized. Systemic anticoagulation and surgical thrombectomy constituted the mainstay treatment. This represents the first report of such devastating cardiovascular complications after anabolic steroid abuse and their management.

Mid-term results with 1,503 CarboMedics mechanical valve implants.

BACKGROUND AND AIM OF THE STUDY: The CarboMedics bileaflet prosthetic heart valve was introduced in 1986, and first implanted by the authors in March 1991. The aim of this study was to analyze the authors' clinical experience with this valve. METHODS: Between March 1991 and October 1998, 1,503 valves were implanted in 1,350 patients (758 males, 592 female; mean age 62 +/- 13 years). Follow up was 99% complete and totaled 4,342 patient-years (pt-yr). RESULTS: The hospital mortality rate was 4.3% (59/1,350). Preoperative NYHA class (p = 0.012), emergency surgery (p = 0.03) and cardiopulmonary bypass time (p = 0.01) were significantly associated with increased risk of operative death (multiple logistic regression). Mean (+/- SEM) survival rates at one and five years were 92.0 +/- 0.7% (n = 1,109) and 80.0 +/- 1.3% (n = 335). Freedom from valve-related complications (linearized rate 5.6%/pt-yr) at one and five years was 89.5 +/- 0.8% (n = 1,031) and 76.3 +/- 1.4% (n = 284). Linearized rates for bleeding events were 2.19%/pt-yr, thromboembolic events 2%/pt-yr, operated valvular endocarditis 0.18%/pt-yr, valve thrombosis 0.14%/pt-yr and non-structural dysfunction 1.22%/pt-yr. Freedom from reoperation at one and five years was 98.5 +/- 0.3% (n = 1,107) and 97.3 +/- 0.5% (n = 334). CONCLUSION: Mid-term results demonstrate that the CarboMedics prosthetic heart valve exhibits a low incidence of valve-related complications.

Aortic valve replacement in children: are mechanical prostheses a good option?

OBJECTIVE: The choice of the most appropriate substitute in children with irreparable aortic valve lesions remains controversial. The aim of this study was to assess early and late outcomes following aortic valve replacement (AVR) with mechanical prostheses in children. PATIENTS: Fifty-six patients (42 male, 14 female, mean age 11.2, range 1-16 years) undergoing AVR with mechanical prostheses between October 1972 and January 1999 were evaluated. Thirty-six patients (64.2%) underwent previous cardiac surgery. Disease aetiology was congenital in 47 patients (congenital aortic stenosis in 33, and other congenital abnormalities in 14) (83.9%), infective in four (7. 1%), rheumatic in two (3.4%), and three (5.3%) had connective tissue disorders. Haemodynamic indication for AVR was aortic regurgitation (AR) in 24 (42.8%), aortic stenosis (AS) in 22 (39.2%) and mixed disease in ten (17.8%). Twenty-eight patients (50.0%) were in New York Heart Association (NYHA) class III-IV before surgery. Concomitant procedures were performed in 31 patients (55.3%), including aortic root enlargement in 28 (50%). The mean size of implanted valves was 22.4 mm (range 17-27 mm). All patients received long-term anticoagulation treatment with sodium warfarin, aiming to maintain an international normalized ratio (INR) between 2.5-3.0. The mean follow-up was 7.3 years (range 0-26, total 405 patient-years). RESULTS: Operative mortality was 5.3% (three patients). Three patients developed complete heart block requiring pacing, two of them permanently. Late events included valve thrombosis (one), transient stroke (one), paravalvular leak of a mitral prosthesis (one), aneurysm of sinus of Valsalva (one) and pannus ingrowth (one). There was no major haemorrhagic event. Five patients required re-operation (8.9%), but none due to outgrowth of the valve. Regarding actuarial freedom from thrombo-embolism, any valve-related event and re-operation at 20 years was 93, 86.6 and 86. 4%. There were three late deaths. Actuarial survival, including operative mortality, at 10 and 20 years was 91 and 84.9%. The actuarial survival for the group of the patients with congenital AS (n=33) at 10 and 20 years was 93.5%, whereas for the children with other congenital heart problems (n=14) this was 85.7 and 64.3% (P=0. 09). At the latest clinical evaluation, 44 children were in NYHA class I and six were in class II. The mean gradient across the aortic prosthetic valve on echocardiography was 17.9 mmHg (range 0-47 mmHg). CONCLUSIONS: Mechanical AVR, with enlargement of the aortic root if necessary, remains an excellent treatment option in children. It is associated with acceptable operative mortality, low incidence of late events and re-operation, and provides good long-term survival. It clearly represents a good alternative to available biological substitutes, including the pulmonary autograft (Ross procedure).

A single-center experience with 1,378 CarboMedics mechanical valve implants.

BACKGROUND: The CarboMedics bileaflet prosthetic heart valve was introduced in 1986. We first implanted it in March 1991. The purpose of this study was to analyze our clinical experience with this valve. METHODS: Between March 1991 and December 1997, 1,378 valves were implanted in 1,247 patients, 705 men (56.5%) and 542 (43.5%) women with a mean age of 62 +/- 11.9 years (+/- the standard deviation). Follow-up is 99% complete and totals 3,978 patient-years. RESULTS: The early mortality rate was 4.4% (55/1,247). The survival rates at 1 year and 5 years were 91.8% +/- 0.8% (+/- the standard error of mean) (n = 1,062) and 79.2% +/- 1.4% (n = 281), respectively. Freedom from valve-related complications (linearized rate, 4.9% per patient-year) at 1 year and 5 years was 90.6% +/- 0.8% (+/- the standard error of the mean) (n = 996) and 80.6% +/- 1.4% (n = 243), respectively. Linearized rates for various complications were as follows: bleeding events, 1.73% per patient-year; embolic events, 1.76% per patient-year; operated valvular endocarditis, 0.18% per patient-year; valve thrombosis, 0.10% per patient year; and nonstructural dysfunction, 1.21% per patient-year. Freedom from reoperation at 1 year and 5 years was 98.6% +/- 0.3% (+/- the standard error of the mean) (n = 1,070) and 97.7% +/- 0.5% (n = 285), respectively. CONCLUSIONS: Midterm results demonstrate that the CarboMedics prosthetic heart valve exhibits a low incidence of valve-related complications.

Replacement of the proximal aorta and aortic valve using a composite bileaflet prosthesis and gelatin-impregnated polyester graft (Carbo-Seal): early results in 143 patients.

OBJECTIVE: We report the combined early results from two centers in the United Kingdom using a composite conduit consisting of a bileaflet mechanical valve incorporated into a gelatin-impregnated, ultra-low porosity, woven polyester graft (Carbo-Seal; Sulzer Carbomedics, Inc, Austin, Tex). METHODS: Between August 1992 and March 1997, 143 patients underwent aortic root replacement with the Carbo-Seal composite prosthesis. The indication for surgery was acute type A dissection in 31 (22%), chronic type A dissection in 9 (6%), ascending aortic aneurysm without dissection in 100 (70%), and false aneurysm of the ascending aorta in 3 (2%). Twenty-seven patients (19%) had undergone previous sternotomy, and 40 (28%) were seen as emergencies. Concomitant procedures were performed in 38 (27%), including 18 aortic arch or hemiarch replacements. Total follow-up is 270 patient-years. Follow-up is 100% complete. RESULTS: The early (30-day) mortality was 7% (10 patients). Permanent neurologic events occurred in 2%. At a mean follow-up of 23 months, 94% of survivors were in New York Heart Association functional class I. Freedom from reoperation was 97.2% +/- 1.6% (1 standard error [1 SE]) at 12 months and 95.7% +/- 2.2% at 48 months. Including early mortality, survival was 90.1% +/- 2.6% at 12 months and 83.1% +/- 3. 5% at 48 months. CONCLUSIONS: Aortic root replacement with use of the Carbo-Seal prosthesis can be undertaken with a relatively low early mortality and morbidity. A low reoperation rate and high intermediate-term survival can be expected, but continued follow-up is needed to determine the long-term efficacy of this prosthesis.

Lung resection for the treatment of severe localised bronchiectasis in cystic fibrosis patients.

BACKGROUND: A small proportion of cystic fibrosis patients develop severe localised bronchiectasis. When this persists despite maximal medical therapy it presents a difficult management problem. Lung transplantation cannot be justified. We report encouraging results in six patients with severe localised bronchiectasis and cystic fibrosis who underwent pulmonary resection. METHODS: Each child had severe localised bronchiectasis despite maximal medical therapy. Intensive preoperative toilet was instituted and pulmonary resection undertaken when lung function was optimal. RESULTS: There was a marked improvement in symptoms in every case. No significant long-standing morbidity was associated with the resection. There was no significant decrease in pulmonary function following resection. CONCLUSION: Pulmonary resection should be considered in the management of severe localised bronchiectasis unresponsive to maximal medical therapy in cystic fibrosis patients.

Surgery for oesophageal carcinoma in the elderly.

OBJECTIVE: There has been a gradual increase in the number of elderly patients referred for oesophageal surgery. The aim of this study is to review our experience with oesophageal cancer surgery in the elderly. METHODS: Between January 1974 and December 1996, 591 patients (408 males, 183 females; mean age 66 years) underwent an oesophageal resection for carcinoma. 221 were aged greater than 70 years of age (group A) and 370 less than 70 (group B). RESULTS: Total in hospital mortality was 8.8% (52/591). This has decreased to less than 5% over the last decade. There was no significant difference in perioperative morbidity or mortality between the groups (P = 0.11). When deaths from unrelated medical conditions were excluded, there was no significant difference in survival between the different age groups (P = 0.96). CONCLUSION: Oesophageal surgery can be performed in a selected elderly population with a low operative morbidity and mortality. The survival benefit of resection is the same in the elderly as for younger patients.

Autotransfusion of washed shed mediastinal fluid decreases the requirement for autologous blood transfusion following cardiac surgery: a prospective randomized trial.

OBJECTIVES: The National Blood Service issues 2.2 million units of blood per year, 10% of these (220000) are utilized in cardiac procedures. Transfusion reactions, infection risk and cost should stimulate us to decrease this transfusion rate. We test the efficacy of autotransfusion following surgery in a prospective randomized trial. METHODS: One hundred and twelve patients undergoing CABG, valve or CABG + valve procedures were randomized into two groups. Group A received washed postoperative drainage fluid and group C were controls. The indication for transfusion was a postoperative haemoglobin (Hb) < 10 g/l or a PCV < 30. There was no significant difference in preoperative and operative variables between the groups. RESULTS: Twenty-eight patients in group A and 46 in group C required homologous transfusion (P = 0.0008). Group A patients required 298+/-49 ml of banked blood per patient, group C 508+/-49 ml (P = 0.003). There was no difference in total blood required (volume autotransfused + volume banked blood transfused) between the groups (group A 404+/-50 ml, group C 508+/-50 ml) or in mean total mediastinal fluid drainage (group A 652+/-51 ml, group C 686+/-50ml). The mean Hb concentration was significantly higher in group A on day 1 (11.2 g/dl+/-51 vs. 10.6 g/dl+/-13 (P = 0.002)). No morbidity was associated with autotransfusion. CONCLUSION: Autotransfusion can decrease the amount of homologous blood transfused following cardiac surgery. This represents a benefit to the patient and a decrease in cost to the health service.

Oesophagectomy for carcinoma of the oesophagus and oesophagogastric junction.

OBJECTIVE: Oesophageal carcinoma has a poor prognosis; surgical resection remains the only chance of cure but is still associated with a significant morbidity and mortality. The aim of this study was to review the results of one surgeon for oesophageal resection for carcinoma of the oesophagus and oesophagogastric junction over a 23 year period. METHODS: Between January 1974 and December 1996, 591 patients (408 males; 183 females; mean age 66 years) underwent an oesophageal resection for carcinoma of the oesophagus or oesophagogastric junction. RESULTS: In hospital mortality was 8.8% (52/591). This has decreased to less than 5% for resections between 1985 and 1996. Non-fatal complications occurred in 21% of patients (123/591). Survival, including in hospital mortality (+/-SEM), was 53.98% (+/-2), 31.77% (+/-2) and 15.3% (+/-2) at 1, 2 and 5 years respectively. CONCLUSION: Early mortality following oesophageal resection has fallen in recent years. Despite considerable experience, long term survival remains disappointingly low.