Light at the end of the tunnel? Palliation for oesophageal carcinoma.

Palliation of incurable oesophageal carcinoma is important, because radical ablation carries a high mortality and because the unpleasant alternative of terminal complete oesophageal obstruction needs to be avoided. The ideal palliative technique would provide normal swallowing for the patient's remaining days by a technique which was quick, safe, painless, needing only a short inpatient stay and having a low complication rate. Attempts to secure this paradigm have resulted in many treatments, few of which have been subjected to good controlled trials.' 2 The three main methods of palliation are surgery, intubation and now laser therapy: two papers published in this issue of Gut"4 bring encouraging results of the last. Operation by a skilled surgeon can in selected cases lead to excellent palliation because it usually relieves dysphagia completely. The considerable morbidity and a mortality of 7 to 29% even in skilled hands, however, is a high price to pay.5' Operations designed to avoid thoracotomy, intrathoracic anastomosis, multiple intra-abdominal anastomoses and a bypassed unresected oesophagus probably give the best and safest surgical palliation, although Ellis has reported a series of 167 patients treated by oesophagogastrectomy with only two deaths and 22 major complications." While age itself is not a contraindication to surgery,'2-'4 many older patients may prefer to tolerate less than perfect swallowing in order to avoid surgery. Discarding the surgical option may mean that a few patients miss an unrecognised chance of cure, because pre-operative staging of oesophageal carcinoma is notoriously difficult, even with computerised tomography and endoscopic ultrasound. Simple bougie or balloon dilatation of the malignant stricture still has a place not all growths narrow down quickly and in patients with a very short life expectancy one or two bougienage sessions may suffice. In others a trial of bougienage is worthwhile as a prelude to endoscopic intubation. Multiple dilatations are usually needed, however,'5 6 and one should not set out to bring the patient back repeatedly to hospital for such therapy. Tubes placed through the growth at the time of surgical exploration have a long history, but planned surgical intubation through a gastrotomy in patients known to have inoperable growths has been generally abandoned because of the disproportionately high morbidity and mortality.5 17 Endoscopic intubation has many attractive features, especially for the frail and elderly. The single procedure, usually without general anaesthesia, short hospital stay and an immediate improvement in swallowing are considerable gains. The 'tube existence' which forbids solid food, the care required to keep the tube patent and the risk of subsequent tube migration, however, detract from it. The operator's expertise is vital and frail patients