We aim to compare the outcomes of patients undergoing R0 esophagectomy by a multidisciplinary team (MDT) with outcomes after surgery alone performed by surgeons working independently in a UK cancer unit. An historical control group of 77 consecutive patients diagnosed with esophageal cancer and undergoing surgery with curative intent by six general surgeons between 1991 and 1997 (54 R0 esophagectomies) were compared with a group of 67 consecutive patients managed by the MDT between 1998 and 2003 (53 R0 esophagectomies, 26 patients received multimodal therapy). The proportion of patients undergoing open and closed laparotomy and thoracotomy decreased from 21% and 5%, respectively, in control patients, to 13% and 0% in MDT patients (chi2 = 11.90, DF = 1, P = 0.001; chi2 = 5.45, DF = 1, P = 0.02 respectively). MDT patients had lower operative mortality (5.7%vs. 26%; chi2 = 8.22, DF = 1, P = 0.004) than control patients, and were more likely to survive 5 years (52%vs. 10%, chi2 = 15.05, P = 0.0001). In a multivariate analysis, MDT management (HR = 0.337, 95% CI = 0.201-0.564, P < 0.001), lymph node metastases (HR = 1.728, 95% CI = 1.070-2.792, P = 0.025), and American Society of Anesthesiologists grade (HR = 2.207, 95% CI = 1.412-3.450, P = 0.001) were independently associated with duration of survival. Multidisciplinary team management and surgical subspecialization improved outcomes after surgery significantly for patients diagnosed with esophageal cancer.

Two cases of cystic arterial disease of the popliteal artery are described in which a communication with the capsule of the knee‐joint was demonstrated. The importance of this finding in the aetiology of the condition is discussed.

Modified D2 gastrectomy without pancreatico-splenectomy, performed by specialist surgeons, can improve survival after R0 resections without increasing operative morbidity and mortality, when compared with D1 gastrectomy performed by general surgeons.

SUMMARY The small bowel of 44 rats has been made ischaemic for 120 minutes by occluding the superior mesenteric artery. Gaseous oxygen was introduced into the small bowel lumen of 18 of the rats, and gaseous nitrogen into the bowel lumen of a further eight, 30 minutes after the start of the ischaemia. The mortality rate of thecontrolgroup of 18 rats was 89 %, significantlyhigher than the 39 % mortality of the oxygen-treated rats (p < 0 005). None of the eight rats treated with intraluminal nitrogen survived. Histological changes caused by ischaemia were much reduced in the intestine of the oxygen treated rats.

SUMMARY Absorption of bacterial endotoxin has been shown to occur after release of the occluded superior mesenteric artery in the rat. A significant reduction in endotoxin absorption was observed as a result of the infusion of gaseous oxygen into the bowel lumen during the period of ischaemia (p < 0 001). The methods of endotoxin identification and assay are described, and the possible reasons for this effect of intraluminal oxygen are discussed.The role of bacterial endotoxins in the genesis of the circulatory collapse associated with acute ischaemia of the intestine is still the subject of debate. Some authors (Zweifach et al., 1958;Williams et al., 1968) deny the importance of endotoxaemia in this condition, but the majority consider that significant amounts of endotoxin reach the circulation, because of increased absorption through the damaged mucosa (Nelson and Noyes, 1952;Kobold and Thal, 1963;Milliken et al., 1965).The reduction in mortality resulting from the introduction of oxygen into the bowel lumen of rats subjected to superior mesenteric artery occlusion has been described previously (Shute, 1975). In that study two identical groups of rats were subjected to occlusion of the superior mesenteric artery (SMA) under general anaesthetic. Thirty minutes after starting arterial occlusion 20 ml of gaseous oxygen was introduced into the small bowel lumen of each of one group of rats, while the other group acted as controls. The SMA clamp was released in all animals after a total of 120 minutes occlusion and the mortality was observed 48 hours later. The mortality of the control group at this time was 89%, while that of the oxygen treated group was 39%. This difference was highly significant P < 0-005 (chi square test).The present study was designed to evaluate the effect of the administration of intraluminal oxygen on the absorption of endotoxin after experimental occlusion of the superior mesenteric artery in the rat. Endotoxin absorption was assessed by the method of Received for publication 30 November 1976 Pieroni et al. (1970), who showed that submicrogram quantities of endotoxin could be detected by assessing the mortality after simultaneous intraperitoneal injection of the endotoxin and 12 5 ug actinomycin D into mice, because the latter drug enhances the lethal effect of endotoxin by up to 225 000 times. MethodsENDOTOXIN ASSAY (Table 1) To confirm the efficacy of the Pieroni endotoxin assay, the following experiment was carried out.Twenty female Swiss A2G mice weighing 20-25 g were injected with 1 ml normal saline and 12-5 ,tg actinomycin D in 0-25 ml saline. Forty similar mice were injected with 0 005 ,ug Escherichia coli 055 :B5 lipopolysaccharide/endotoxin (Difco, England) in 0 5 ml saline, and a further 50 mice received the same amounts of actinomycin D and endotoxin simultaneously. All injections were given intraperitoneally, and the ampoules of actinomycin D were pooled to avoid the problem of variation in potency between