Authors writing an oesophageal cancer include adenocarcinoma to a variable extent--between 1 and 75 per cent--but the true incidence of this histological type is about 1 per cent. Most adenocarcinomas are gastric in origin, involving the lower oesophagus, have a lower operative mortality than in the middle or upper one-third of the oesophagus and poorer prognosis than squamous cell carcinoma, but there is no alternative treatment to surgery. Squamous cell carcinoma of the oesophagus, separated incompletely but as far as possible, has been analysed by reviewing data on 83 783 patients in 122 paERS. After trying to standardize the data, it appears that of 100 patients with the condition, 58 will be explored and 39 have the tumour resected, of whom 13 will die in hospital. Of the 26 patients leaving hospital with the tumour excised, 18 will survive for 1 year, 9 for 2 years and 4 for 5 years. Oesophageal resection for squamous cell carcinoma has the highest operative mortality of any routinely performed surgical procedure today.

There has been no controlled trial of radiotherapy versus surgery for squamous cell carcinoma of the oesophagus. Radiotherapy is generally used for those patients with extensive disease or those who are unfit for surgery. In spite of this, the 1-year survival of 18 per cent is similar to that for surgically treated patients and there is no equivalent operative mortality. The 5-year survival is 6 per cent compared with that for surgery of 4 per cent. There are no results available to suggest what would happen if a patient with a localized tumour, technically suitable for surgical resection, were treated instead by radiotherapy.

Literature review (1970-85) indicates excellent or good results following Heller's operation in 89 per cent of 5002 patients. The overall mortality was 2.8 per cent with a reoperation rate of 2.8 per cent. When the operation was done through an abdominal incision, gastro-oesophageal reflux was almost twice as common as when it was done through a thoracic incision, regardless of whether an anti-reflux procedure was performed.

A review of the different bougies used over the years confirms that the dilators presently available are adequate for the conservative management of oesophageal benign strictures. Most of these are of an old but well-tested design. Several additional techniques may have to be tried in difficult patients. Dysphagia can be relieved by the passage os size 39 FG bougies but many will wish to increase this to 60 FG, if it is possible. Failure of conservative management by dilatation only can be defined as (a) technical impossibility to dilate sufficiently to relieve the dysphagia, which is rare, or (b) when the patient or the surgeon considers the procedure is being done too frequently, which must be measured in FG for the bougie and weeks for frequency. The more expert the surgeon becomes at dilatation, the safer it will become, and the necessity for surgical intervention will be less frequent.

SUMMARY Responses of the cricopharyngeal sphincter to graded intraluminal distension were studied in order to determine its response threshold and to define the functional relationship between the sphincter and oesophageal body. Nine normal subjects underwent manometric study using a multilumen tube with an attached inflatable balloon sited 10 cm below the sphincter. Sphincteric and oesophageal motor responses to six graded balloon inflations were recorded in each subject. The sphincter responded to distension with increasing rise in pressure, from a median value of 42-5 mmHg at lowest levels of distension to 95 mmHg at maximal tolerated distension. Non-swallow related contractile activity was stimulated in the oesophageal body proximal to the distension and increased in quantity as inflation progressed. Distal propagation of this secondary activity was progressively inhibited with increasing distension. These interrelated changes thus show the normal upper oesophageal clearance responses to intraluminal distension. It is suggested that their more widespread application, in addition to standard manometric techniques, might provide a more rational evaluation of those patients suspected to have impaired oesophageal clearance, but in whom standard manometry is non-diagnostic.The cricopharyngeal (upper oesophageal) sphincter, situated at the intersection of the airway and the first part of the alimentary tract' is formed mainly by the cricopharyngeal muscle, with additional fibres from the circular muscle of the oesophagus distally and the inferior pharyngeal constrictor proximally.23The coordinated relaxation and contraction of this sphincter constitutes an integral part of normal deglutition, ensuring the passage of a bolus from the pharynx into the oesophagus, and additionally forms a dynamic barrier to prevent oesophagopharyngeal reflux and spillover into the tracheobronchial tree. This latter function is of particular importance because breakdown of this mechanism may be related to oesophagopharyngeal regurgitation and aspiration pneumonitis. '34 Measurements of the resting sphincter pressure and its relationships to the oesophageal body have previously been attempted,>' but the data are incom-