Fat malabsorption in patients before and after total gastrectomy, studied by the triolein breath test.

The feasibility of using the triolein breath test to demonstrate fat malabsorption was evaluated in a prospective study of patients before and after total gastrectomy and Roux-en-Y reconstruction. Two of 11 patients had subnormal fat absorption before the operation, but 1 and 6 months after the operation 9 of 11 patients had subnormal fat absorption. Peak expiratory 14CO2 (median (range] at the three investigations was 3.9%/h (2.1-5.9%/h), 2.1%/h (1.4-4.5%/h), and 2.0%/h (1.2-6.0%/h), respectively. Patients who underwent a Nissen fundoplication were used as controls. They had normal fat absorption both before and after operation. Serum amylase was not appreciably affected by total gastrectomy and was similar to control values. In contrast, serum albumin decreased 1 month after gastrectomy and recovered after 6 months. In control patients pre- and post-operative albumin concentrations did not differ and were comparable to preoperative albumin values in the total gastrectomy group. The lowered fat absorption may be explained by duodenal bypass with decreased pancreatic stimulation, and it may in part explain the weight loss in patients operated on with total gastrectomy.

Influence of aspirin and steroids on acute lung injury after i.v. Injection of a sclerosing agent.

Recently it was reported that sclerosing agents frequently used for sclerotherapy of bleeding oesophageal varices can cause severe lung injury, resembling that of the adult respiratory distress syndrome (ARDS), if given intravenously to sheep. The aim of the present series was to study if steroids or aspirin given before injection of the sclerosing agent would modify or prevent the lung injury. Twenty-one sheep were randomly given either normal saline (S, 20 ml i.v., n = 8), methyl prednisolone (MP, 40 mg/kg bwt i.v., n = 6) or aspirin (ASP, 10 mg/kg bwt i.v., n = 7) 20 min before ethanolamine oleate 5% 5 ml i.v. was given. After injection of the sclerosing agent the untreated animals (S) developed a high pulmonary artery pressure (PAP) while the respiratory compliance (CT), arterial oxygen tension (PaO2) as well as the platelet (PC) and white cell counts (WBC) decreased significantly. In group MP the pulmonary changes were less severe (CT, PaO2) and the drop in PC as well as WBC was less pronounced, but the PAP increased just as much as in the controls (S). In the aspirin treated group (ASP), however, no pulmonary or circulatory changes were observed. PAP, PaO2, CT, PC and WBC remained virtually unchanged throughout the study. At post mortem the lungs of the control animals appeared moderately congested and the wet/dry weight ratio was significantly increased compared with group ASP. It was concluded that methyl prednisolone slightly attenuates acute lung injury observed after i.v. injection of the sclerosing agent ethanolamine oleate. Aspirin, on the other hand, appears to prevent this injury almost entirely.

Lack of correlation between fecal blood loss and drug-induced gastric mucosal lesions.

Increased fecal blood loss was produced in healthy volunteers by the administration of two nonsteroidal anti-inflammatory drugs (NSAID), naproxen or fenflumizole. Basal as well as drug-induced gastrointestinal blood loss was measured using 51Cr erythrocyte labeling. Median rise in daily fecal blood loss was 432%. All subjects were endoscoped at the initiation and at the completion of the study. Endoscopic findings were assessed quantitatively by two observers in two different ways. All subjects but three had gastric mucosal lesions at follow-up endoscopy. There was a good correlation between the endoscopic assessments but no statistical correlation between the endoscopic assessment and the increase in fecal blood loss. The data suggest that factors other than gastric mucosal lesions have to be taken into account when investigating NSAID-induced gastrointestinal bleeding.

Evaluation of 99Tcm mercaptoacetyltriglycine for the detection and localization of gastrointestinal bleeding in an experimental animal model.

Technetium-99m mercaptoacetyltriglycine (99Tcm-MAG3) was evaluated for the detection and localization of the site of gastrointestinal bleeding in a sheep model. Radioactivity was detected in both the stomach and lower abdomen. However, 99Tcm-MAG3 is partially excreted by the liver and in the bile. The radioactivity in the bile moved to the small bowel. This movement of radioactivity in the lower abdomen can lead to a misinterpretation of the site of bleeding. Hence, 99Tcm-MAG3 may not be an effective radiopharmaceutical for the localization of the site of bleeding in the gastrointestinal tract, particularly the lower abdomen.

Influence of intravenous injection of sclerosing agents on the respiratory function.

Severe lung injury (ARDS) has occasionally been observed after sclerotherapy for bleeding oesophageal varices. In order to study the effects of sclerosing agents, which may escape into the systemic circulation during treatment, seven sheep were given either ethanolamine oleate (n = 4) or sodium tetradecyl sulfate (n = 3) intravenously. A control group (n = 3) was treated identically with the others except it did not receive any sclerosing agent. The study showed that both sclerosing agents caused an immediate and severe fall in total respiratory compliance and arterial oxygen tension. There was a marked trapping of platelets in the lungs, which was also reflected by a drop in platelet count in peripheral blood. The lungs from the animals receiving sclerosing agents appeared moderately to severely congested and the wet/dry weight ratio of the lungs was significantly increased compared with untreated normal lungs (p less than 0.01). Histopathological examination revealed severe damage to the alveolar membranes, intraalveolar fibrino-haemorrhagic exudate, collapse of alveolar spaces and numerous eosinophilic leukocytes in the broadened, oedematous alveolar walls. It was concluded that the sclerosing agents used in this study, ethanolamine oleate and sodium tetradecyl sulphate, cause severe lung injury if given intravenously in sheep in doses corresponding to 25-50% of what is normally used during sclerotherapy in patients. The mechanism of this action may be that of an increased microvascular permeability causing marked alveolar damage and destruction of the blood gas barrier of the lungs.

Comparison of four technetium-99m radiopharmaceuticals for detection and localization of gastrointestinal bleeding in a sheep model.

Four Tc-99 radiopharmaceuticals, Tc-99m sulphur colloid, Tc-99m red blood cells (RBCs), Tc-99m mercaptoacetyltriglycine (MAG3), and Tc-99m DTPA, were studied in an experimental animal model for detection and localization of gastrointestinal (GI) bleeding site in both the upper and lower abdomen. With Tc-99m sulphur colloid and Tc-99m RBCs, it was possible to detect and localize the GI bleeding site in the lower abdomen. With Tc-99m MAG3, it was possible to visualize the bleeding site in both the upper and lower abdomen. However, Tc-99m MAG3 is partially excreted by the liver into the bile, hence it will be difficult to use Tc-99m MAG3 to localize the GI bleeding site in the lower abdomen. With Tc-99m DTPA, it was possible to detect and localize the GI bleeding site simultaneously in both upper and lower abdomen. The overall background radioactivity was reduced considerably by diuresis with frusemide and catheterization of the urinary bladder.

Acid reduction by intragastric contact Nd:YAG laser photoradiation in rats.

Gastric acid reduction in the treatment of duodenal ulcers occurs following extragastric vagotomy, antrectomy, or seromyotomy. This study examines the effects of intraluminal LCV and/or pyloric MA using laser photoradiation in Sprague-Dawley rats. Groups were compared to sham-operation and truncal vagotomy with pyloroplasty (TVP) regarding gastric acid secretion, plasma gastrin, pancreatic polypeptide, and histology, including immunocytochemistry for G-cells. All operative procedures reduced acid output (P less than 0.001) both immediately and at 8 weeks, with LCV being more effective than TVP or MA. Moreover, LCV with MA was no more beneficial than LCV alone. Plasma gastrin increased slightly following TVP but not following the LCV group in the acute study, and no difference was seen in the chronic group. Histologically, the depth of laser penetration was optimum with no perforations. Although the G-cells were destroyed following the laser photoradiation, they reappeared in lesser numbers at 8 weeks. Laser photoradiation of the intramural nerve plexuses on the lesser curve of the stomach by intragastric mucosal myotomy is effective in reducing gastric acid and may have a clinical and therapeutic role with flexible fiberoptic gastroscopy.

Effects of cimetidine, atropine and pirenzepine on basal and stimulated gastric acid secretion in the rat.

The gastric anti-secretagogue effects of cimetidine (a histamine H2-receptor antagonist) and of atropine (a non-selective muscarinic receptor antagonist) and pirenzepine (a selective muscarinic M1-receptor antagonist) were examined in conscious gastric fistula rats both under basal conditions and after stimulation with maximal doses of pentagastrin and histamine. Cimetidine blocked basal as well as stimulated acid secretion. The cimetidine dose-response curves and the calculated ED50 values were similar in the different experimental situations. Atropine blocked equally effectively the basal and the stimulated acid secretion. The antisecretagogue and pupil dilating effects were compared. The ED50 values for the anti-secretagogue effect and for the pupil dilating effect were in the same range though not identical. Pirenzepine blocked acid secretion, whether basal or stimulated, with similar potency. It was much more potent to block acid secretion than to cause pupil dilatation. The greater potency of pirenzepine to block acid secretion than to cause pupil dilatation suggests that the cholinergic pathway of acid secretion involves neuronal muscarinic M1-receptors within the intramural ganglia of the stomach wall. In conclusion, cimetidine, atropine and pirenzepine effectively blocked basal as well as pentagastrin- and histamine-stimulated acid secretion, indicating that both histamine and acetylcholine are important in the control of the parietal cell. Histamine has been claimed to be the final common chemical mediator of acid secretion. This view is at odds with the fact that muscarinic blocking agents also inhibit basal and stimulated acid secretion.

Esophagojejunostomy with the EEA stapler.

Esophagojejunostomy after total gastrectomy was attempted in 27 operations with the EEA stapling device (U.S. Surgical Corp., Norwalk, Conn.). After removal of the specimen the anastomosis is performed with an end-to-side technique with insertion of the cartridge and its central rod through the open jejunal end. The 28 mm wide cartridge was used in 24 anastomoses and the 25 mm wide cartridge was used in two. In one case the 25 mm cartridge tore the distal esophagus, and the anastomosis had to be sutured manually. The median operation time was 305 minutes (range, 205 to 560 minutes), and the time to perform the anastomosis was 20 minutes (range, 15 to 60 minutes). Anastomotic leakage occurred in three patients, two of whom were stapled with the 25 mm cartridge. All healed with conservative treatment. One patient developed a stricture at the anastomotic site due to recurrence of the tumor. There was one hospital death. Median hospital stay was 16 days (range, 8 to 71 days) and median survival time was 11 months. It is concluded that the EEA stapler allows the construction of a fast and reliable esophagojejunostomy with good functional results after total gastrectomy for gastric cancer.

Effect of Intralipid on clearance and organ uptake of bacteria in intra-abdominal sepsis in rats.

Although infusion of Intralipid has been reported to block reticuloendothelial system (RES) function, it is unclear if this is true at rates used clinically. This study investigates the effect of Intralipid, infused at a rate providing about half of the non-protein caloric needs, on survival and bacterial clearance in septic rats. Continuous infusion of Intralipid (0.9 g/100 g bw/d) or saline (controls) was started immediately after induction of hyperdynamic intra-abdominal sepsis (bacterial agents: E. coli, B. thetaiotamicron). RES function was studied by means of intravenous injection of Selenium-labelled, viable E. coli after 24 or 48 h. Compared to the saline-treated controls, Intralipid did not cause any change in clearance from blood or localisation to liver, spleen or lungs. The 24 and 48 h survival rates were about 80 and 65%, respectively, and similar in the two groups. It is concluded that infusion of Intralipid at a rate close to that used clinically did not impair survival or bacterial clearance in rats with gram-negative sepsis.

Activation and hyperplasia of gastrin and enterochromaffin-like cells in the stomach.

The stomach is rich in endocrine cells, most of which are still unidentified with respect to the peptide hormones they produce. The endocrine cell populations in the antrum usually differ from those in the oxyntic mucosa. Gastrin cells are found in the antrum and respond readily to stimuli from the gastric lumen, such as changes in the pH and the presence of food. In order to study the functional control of the antral gastrin cell, rats were subjected to different kinds of surgery. The serum gastrin concentrations in the various experimental groups were measured 8-10 weeks after the operations. Elevated antral pH raised the serum gastrin concentration. The combination of elevated antral pH and the passage of food over the pyloric glands produced gastrin cell hyperplasia. The operation that was most effective in inducing gastrin cell hyperplasia was removal of the acid-producing part of the stomach. Interestingly, gastrin cell hyperplasia was seen also after bilateral truncal vagotomy, indicating that an intact vagal innervation is not essential for the development of gastrin cell hyperplasia. Enterochromaffin-like (ECL) cells are endocrine/paracrine cells that are numerous in the acid-producing part of the stomach in many species. In the rat, they occur predominantly in the basal half of the oxyntic mucosa and produce and store histamine. The ECL cells have an unknown function and do not seem to respond to stimuli from the gastric lumen. They are activated by circulating gastrin and by vagal excitation. Gastrin mobilises histamine from these cells and activates the histamine-forming enzyme, histidine decarboxylase. Long-term hypergastrinaemia produces diffuse ECL cell hyperplasia, whereas hypogastrinaemia (following removal of the endogenous stores of gastrin by antrectomy) reduces the ECL cell number. Portacaval shunt brings about a marked increase in the number of ECL cells through an unknown mechanism. Also neuronal stimuli are important for the trophic control of the ECL cells. Studies of unilaterally vagotomised rats showed reduced weight and thickness of the oxyntic mucosa as well as a markedly reduced number of ECL cells on the denervated side. Gastric carcinoids in man are rare tumours predominantly made up of ECL cells. The incidence of such tumours is increased in patients with hypergastrinaemia (pernicious anaemia, Zollinger-Ellison syndrome). A diffuse ECL cell hyperplasia is a common finding in such patients, which is in keeping with the known gastrin sensitivity of the normal ECL cell in the rat.

Branched chain amino acid-enriched solutions in the septic patient. A randomized, prospective trial.

A prospective, randomized trial was undertaken to compare the nutritional efficacy in surgical stress of a standard amino acid solution and two branched chain-enriched amino acid solutions: one enriched primarily with valine, the other with leucine. The study comprised 37 patients in the surgical intensive care unit who received isocaloric, isonitrogenous parenteral nutrition started within 24 hours of the onset of major operation, injury, or sepsis. Nitrogen retention was marginally but statistically significantly better on days 5, 7, and 10 in both groups of patients receiving the branched chain-enriched solutions, but differences in cumulative nitrogen balance were not statistically significant. Amino acid composition appeared to be important in that the group receiving the leucine-enriched solution appeared to maintain hepatic protein synthesis better (as manifest by higher short-turnover plasma protein concentrations) and required less exogenous insulin to maintain euglycemia. Improved outcome was not seen in the groups receiving the branched chain-enriched solutions.

Preoperative computed tomography for evaluation of tumour growth in patients with gastric cancer.

In 31 patients with proven gastric malignancy, computed tomography (CT) was done before surgical exploration. Tumour growth was assessed with each method independently, but following the same protocol. The results of the CT examination were compared with the surgical findings and the diagnostic accuracy of CT was evaluated. Liver metastases could be demonstrated with specificity 10/10 and sensitivity 20/21. The predictive value of the positive test (Pv+) was 10/11, and the predictive value of the negative test (Pv-) was 20/20. Tumour growth to the pancreas was shown by CT with specificity 4/5 and sensitivity 26/26, with Pv+ = 4/4 and Pv- = 26/27. Enlarged lymph nodes around the coeliac axis could be shown with specificity 6/6, Pv+ = 18/18 and Pv- = 6/13. The accuracy of CT was also evaluated regarding tumour growth to oesophagus, colon and greater and lesser omenta. Tumour resectability was predicted by the radiologist with the help of CT. The specificity was 26/26 and sensitivity 4/5, Pv+ = 26/27 and Pv- = 4/4. In this study, therefore, CT had high diagnostic accuracy in regard to extent of intraabdominal growth and resectability. In selected patients it may thus be possible, when resection is not feasible, to avoid exploratory laparotomy.

The vagus exerts trophic control of the stomach in the rat.

Bilateral subdiaphragmatic truncal vagotomy results in great functional changes in the stomach although the changes in the gastric mucosal architecture are small. A trophic effect of the vagus on the stomach is revealed after unilateral vagal sectioning, taking advantage of the fact that, in the rat, each vagal trunk innervates only one side of the stomach, and that denervation of one side does not impair the functional capacity of the other. The denervated side of the stomach displayed atrophy that was reflected in reduced weight and height of the oxyntic mucosa and a reduced density of argyrophil cells. The lack of atrophy after bilateral vagotomy can be explained by counteracting forces, in that the subsequent rise in gastrin secretion (due to lack of acid feedback inhibition of gastrin release) probably masks antitrophic effects of the vagotomy per se. Interestingly, the number of somatostatin cells in the oxyntic mucosa was not reduced after unilateral vagotomy, nor was the weight of the antral mucosa or the density of enterochromaffin and gastrin cells in the antrum on the denervated side.

Neuronal histamine in the gut wall releasable by gastrin and cholecystokinin.

Histamine accumulated in the ligated vagus nerve of the rat, both above and below the ligature; maximum accumulation was after 4 h. The finding is suggestive of axonal flow. Further evidence for histamine in peripheral nerves was obtained in experiments showing that the guinea-pig gut wall could be labelled with [3H]histamine. The experiments were carried out with isolated strips of stomach wall and taenia coli. Electrical stimulation released [3H]histamine from these specimens. The release could be blocked by Ca2+-free medium or by tetrodotoxin. The release was unaffected by vagal denervation or chemical sympathectomy (6-hydroxydopamine) but prevented by reserpinization. Gastrin-17 and cholecystokinin-39 released radioactivity by a tetrodotoxin-sensitive mechanism. The possible existence of a gastrin/cholecystokinin-sensitive neuronal pool of histamine in the gut wall offers a new perspective on the postulated role of histamine as a physiological stimulant of gastric acid secretion and might explain why H2-receptor antagonists block gastrin-stimulated acid secretion.

Mechanism of inhibition of gastric acid secretion by vagal denervation in the rat.

Acute bilateral subdiaphragmatic vagotomy in the conscious fistula rat greatly reduced gastric acid secretion, stimulated by the combined intravenous infusion of pentagastrin (10 micrograms/kg/h), histamine dihydrochloride (3 mg/kg/h) and carbachol (50 micrograms/kg/h). The reduction of acid output was immediate (within 15 min after vagotomy). The greatly reduced acid response to these secretagogues persisted for at least 8 weeks after vagal denervation (longest time studied). The sudden and dramatic effect of vagotomy on acid secretion is not related to a possible deficiency of either acetylcholine or histamine at the respective receptor site since the combined infusion of gastrin, histamine and carbachol did not prevent the suppression of acid secretion. Since the decline in acid output following vagal denervation was immediate, it probably reflects a sudden inaccessibility rather than loss of muscarinic or H2-receptors. The acid output obviously depends upon intramural "transducer" systems that respond to and transmit the vagal input. It is likely that the intramural ganglia represent such "transducer" systems. In the absence of a vagal drive these neuronal "transducers" cease to fire and as a result the parietal cells become almost unresponsive to stimuli.

Gastrins and cholecystokinins release acetylcholine but not substance P from neurons in the guinea-pig taenia coli.

Gastrins and cholecystokinins contract the isolated taenia coli of the guinea-pig. Porcine CCK-39 produced the greatest contractile response and human gastrin-17 I and -34 the weakest. Pentagastrin had the highest affinity to the receptors and non-sulphated CCK-8 the lowest. The contractions produced by the CCK peptides were reduced by the neuronal blocker tetrodotoxin and by the muscarinic blocker atropine but not by the substance P antagonist [D-Pro2,D-Trp7,9]SP. It is concluded that gastrin/CCK peptides act directly on smooth muscle cells and that in addition these peptides, notably sulphated forms of CCK, are capable of exciting cholinergic neurons (but not SP neurons) to cause smooth muscle contraction.

Mechanism of gastric acid response to pylorus ligation: effects of nephrectomy.

In the rat nephrectomy raises the serum gastrin concentration but makes the parietal cells refractory to gastrin. Pylorus ligation stimulates the gastric acid output by a long vago-vagal reflex in innervated animals and by an intramural reflex in chronically vagotomized animals. Nephrectomy reduced the acid response to pylorus ligation in vagally intact rats but enhanced it in vagotomized rats. The acid response to pylorus ligation in all the experimental groups was inhibited by a muscarinic blocker, atropine, and by an H2-antagonist, metiamide. The serum gastrin concentration was raised by nephrectomy and by vagal denervation. Histamine mobilization from gastric endocrine cells is reflected in the activity of gastric histidine decarboxylase. The enzyme activity in pylorus-ligated innervated rats was raised by pentagastrin, atropine, and metiamide. In nephrectomized rats the basal enzyme activity was high, and it was raised further, slightly but significantly, by pentagastrin. The basal enzyme activity in pylorus-ligated rats was also quite high after vagotomy, and it was raised further by pentagastrin. After vagotomy + nephrectomy the basal enzyme activity was very high; it was not raised further by pentagastrin. It appears that both vago-vagal and intramural reflexes involve a cholinergic and a histaminergic pathway, that gastrin is not important for the neurally mediated acid response elicited by pylorus ligation, and that the postulated histaminergic pathway does not involve histamine derived from the gastric endocrine-like cells.

Effect of vagotomy on gastric acid secretion in the rat.

The effects of pentagastrin, histamine or feeding on gastric acid secretion were studied in conscious rats with total gastric by-pass, achieved by transection of the cardia and pylorus, followed by an oesophago-duodenostomy. After closure of the cardia, the by-passed stomach was connected to the small intestine through a Roux-en-Y loop. A chronic gastric fistula was fitted into the rumen. Basal acid output was low in chronically vagotomized rats, being 6% of that in the innervated animals. A clear-cut stimulation was observed after both pentagastrin and histamine in innervated as well as denervated rats, although the maximal acid output in the denervated group was less than 10% of that in the innervated group. In previous studies on acid secretion in vagotomized rats with chronic gastric fistulas, neither basal nor stimulated acid secretion could be detected. Apparently, by-passing the stomach eliminates sources of error associated with the conventional gastric fistula technique (for instance, neutralization of acid gastric juice by swallowed saliva or regurgitated duodenal juice). Nonetheless, the greatly reduced acid output following vagotomy indicates that normal basal as well as normal stimulated acid secretion is dependent upon an intact vagus. Pentagastrin- and histamine-stimulated acid secretion was blocked by atropine and cimetidine in both the innervated and denervated rats. Feeding caused a significant inhibition of acid secretion in the by-passed, innervated stomach. In the denervated stomach feeding was without effect. The mechanism behind the postprandial inhibition of acid secretion in the innervated stomach is obscure. Direct vagal inhibition as well as humoral substances, liberated by vagal stimulation or by the presence of food in the intestine, may be responsible.