Endoscopic haemostasis by injection of adrenaline was attempted in 135 consecutive patients with active upper gastrointestinal bleeding. Initial haemostasis was obtained in 127 patients following injection of 5-15 ml 1:10,000 adrenaline; eight patients in whom haemostasis was not achieved underwent immediate laparotomy. There was further haemorrhage in 25 patients, which was successfully treated by further injection of adrenaline in ten. Fifteen patients had major rebleeding requiring emergency surgery. Stepwise logistic regression analysis identified three factors that, taken together, were highly predictive of the need for surgery: pulse rate on admission, the position of the ulcer and whether the patient was obese. A scoring system was derived from the logistic analysis equation that was found to predict correctly the need for emergency surgery in 84 per cent of patients. In patients with a high probability of rebleeding surgery should be considered after initial endoscopic haemostasis and stabilization. In the majority of patients endoscopic treatment alone is sufficient for permanent haemostasis.

Changes in tumour FDG uptake were seen in all tumours after chemotherapy. FDG-PET may have a role to play in the assessment of patients with upper gastrointestinal malignancy receiving chemotherapy.

Decreased tumour [ 18 F]2-fluoro-2-deoxy-D-glucose ( 18 FDG) incorporation is related to response however its significance at the cell level in gastro-oesophageal cancer and how it relates to cell death is unknown. Here human gastric adenocarcinoma (AGS) cells were treated with lethal dose 10 and 50 (LD 10 and LD 50 ), determined by using the MTT assay, of the three drugs, epirubicin, 5-fluorouracil and cisplatin, commonly used in the treatment of patients with gastro-oesophageal cancer. 18 FDG incorporation was determined after 48 and 72 h of treatment with each drug and related to drug-induced changes in glucose transport, hexokinase activity, cell cycle distribution and annexin V-PE binding (a measure of apoptosis). Treatment of cells for 48 and 72 h with LD 50 doses of cisplatin resulted in reductions in 18 FDG incorporation of 27 and 25% respectively and of 5-fluorouracil reduced 18 FDG incorporation by 34 and 33% respectively: epirubicin treatment reduced incorporation by 30 and 69% respectively. Cells that had been treated for 72 h with each drug were incubated in drug-free media for a further 6 days to determine their ability to recover. Comparison of the ability to recover from the chemotherapy agent, with 18 FDG incorporation before the recovery period allowed an assessment of the predictive ability of 18 FDG incorporation. Cells treated with either 5-fluorouracil or cisplatin demonstrated recovery on removal of the drug. In contrast, cells treated with epirubicin did not recover corresponding with the greatest 72 h treatment decrease in 18 FDG incorporation. In contrast to adherent cells treated with cisplatin or 5-fluorouracil, adherent epirubicin-treated cells also exhibited very high levels of apoptosis. Glucose transport was decreased after each treatment whilst hexokinase activity was only decreased after 72 h of treatment with each drug. There was no consistent relationship observed between 18 FDG incorporation and cell cycle distribution. Our results show that at the tumour cell level in gastric tumour cells, decreased 18 FDG incorporation and glucose transport, accompanies therapeutic growth inhibition. 18 FDG incorporation is particularly diminished in cells exhibiting apoptosis.

Over a 30-month period, 53 patients with actively bleeding non-variceal lesions of the oesophagus, stomach or duodenum were treated by endoscopic injection of 1/10,000 adrenaline. Initial haemostasis was obtained in 50 cases, and permanent haemostasis in 44. Emergency surgery for bleeding was required in nine patients overall, and there were four deaths. All lesions requiring surgery were located on the posterior wall of the duodenum or the lesser curve of the stomach, and all but one had evidence of an exposed arterial vessel. Adrenaline injection is an effective, safe and simple method of endoscopic haemostasis.

Malignant disease is often associated with weight loss and malnutrition. Nutritional support is frequently provided to patients with cancer in an attempt to improve nutritional status and reverse weight loss, with the aim of reducing morbidity and mortality rates. This review evaluates the effect of supplemental nutrition on morbidity and mortality in patients with malignancy undergoing treatment with surgery, chemotherapy or radiotherapy. It also assesses the effect nutritional supplementation has on host defence mechanisms and how nutrients affect tumour cell growth. The evidence suggests that perioperative nutritional support, if given for at least 10 days, reduces morbidity and mortality in patients with biochemical evidence of severe malnutrition, manifest as a low serum albumin concentration and excessive weight loss. In contrast, there is no evidence that parenteral nutritional support benefits patients undergoing chemotherapy or radiotherapy, in terms of either an increased tumour response rate or prolongation of survival. Current research on malignant disease is highlighting the role of specific nutrients (amino acids, essential fatty acids and polyribonucleotides) as key regulators of both anticancer host defence mechanisms and the control of nitrogen metabolism and tumour growth. Arginine, essential fatty acids and ribonucleotides have all been demonstrated to stimulate antitumour host defence mechanisms and some also modulate tumour cell metabolism. Dietary manipulation offers exciting possibilities for the innovative management of malignant disease.

Given H E The role of aspiration cytologic examination in the diagnosis of carcinoma of the breast. Surg Gynerol Obstet 1 99 1 ; lobular carcinoma: mammographic findings in a 10-year 1990; 14: 12-23.breast. Surg Gynecol Obstet 1987; 165: 435-41. 172: 290-2.

Aberdeen and *Raigrnore Hospital, Inverness, UK Correspondence to: M r A. Munro, 4th Floor, Raigmore Hospital, lnverness IV2 3UJ, UK Fundoplication is performed by most general surgeons using an abdominal approach, except in patients who are suspected before operation of having a shortened oesophagus. Occasionally after commencing the procedure through the abdomen, the surgeon finds that despite mobilizing the lower oesophagus, insufficient length of intra-abdominal oesophagus is available to perform an adequate fundoplication. The Collis gastroplasty provides a means of effectively lengthening the oesophagus by creating a lesser curve gastric tube', but this modification usually requires extension of the wound into the left chest, which adds appreciably to the magnitude of the procedure. We report a simple technique which allows the Collis gastroplasty to be performed without recourse to thoracotomy, using a linear cutter stapling instrument.

Natural cytotoxicity (natural killer, NK, and lymphokine-activated killer, LAK, activity) was documented in 12 patients with metastatic colorectal cancer, both before and after a 5-day course of continuous therapy with intravenous recombinant interleukin-2 (rIL-2). Treatment induced a substantial increase in circulating CD56+ lymphocytes (pretreatment: 12.1 +/- 6.9%, mean +/- SD; posttreatment: 39.2 +/- 6.9%. Maximal NK cell activity was induced by treatment with rIL-2 but only suboptimal augmentation of LAK cell cytotoxicity was obtained. This study indicates that although continuous infusion of rIL-2 does have a significant effect on natural cytotoxicity, this is suboptimal and further studies are necessary to define the most efficacious immunity-enhancing regimens of therapy, thereby hopefully improving clinical outcome of rIL-2 treatment.