Long-Term Parenteral Nutrition: Problems with Venous Access

Long-term parenteral nutrition requires central venous access, often difficult in patients who have had several central venous catheterizations. Therapy may be complicated by thrombosis and sepsis which may further compromise central access. We report five cases illustrating such difficulties and suggest that these patients be referred early to specialist centres where experienced catheter insertion and management results in a greatly reduced incidence of complications.

Diagnostic and therapeutic push type enteroscopy in clinical use

This study describes small bowel push enteroscopy in routine clinical practice, using a purpose designed instrument (Olympus SIF-10). Fifty six patients had a total of 60 procedures over a two and a half year period. The median (range) depth of small intestine intubated was 45 (15-90) cm. Procedure time varied from 10-45 minutes. Most enteroscopies were performed during routine gastroscopy lists. The technique was comparatively easy for experienced endoscopists to learn. Forty two procedures were for diagnostic purposes. Eleven patients had gastrointestinal bleeding where the source was obscure, or where early investigations had suggested a small bowel source: a specific diagnosis was made in 45% of these cases. Of seven iron deficient anaemic patients using non-steroidal anti-inflammatory drugs (NSAIDs), only one had a lesion detected in the upper small bowel. Nine patients had abnormal small bowel barium studies. Small bowel abnormalities were seen in six cases and were definitively diagnostic in three of these; in three patients the barium study appearances were confirmed as artefact. Fifteen patients were investigated for abdominal symptoms suggesting small bowel obstruction or malabsorption: a diagnosis was made in five cases. Fifteen patients underwent enteroscopy for therapeutic purposes, including successful treatment of difficult enteral feeding problems by nasojejunal tubes or by cutaneous endoscopic jejunostomies, polypectomy for Peutz-Jeghers syndrome, and dilatation of strictures. Additionally, bleeding lesions detected in patients during investigation of anaemia were successfully treated at the time by YAG laser or bipolar diathermy. In conclusion, push enteroscopy is a practical and valuable clinical service, which should probably become available on a subregional basis.

Increased leukotriene B4 release from ileal pouch mucosa in ulcerative colitis compared with familial adenomatous polyposis

Pouchitis may complicate the construction of an ileal pouch after colectomy for ulcerative colitis (UC) but not familial adenomatous polyposis (FAP). To examine whether differences in eicosanoid metabolism might explain why pouchitis is largely confined to UC patients, this study compared arachidonic acid stimulated release of immunoreactive leukotriene B4 (LTB4) and prostaglandin E2 (PGE2) from macroscopically uninflamed pouch mucosal biopsy specimens incubated in vitro from patients with UC and FAP. The study also compared eicosanoid release from inflamed and uninflamed pouches in patients with UC. In uninflamed pouches, median LTB4 release was nearly twice as high in UC as in FAP (p = 0.001), but there was no significant difference in PGE2 production. In UC, stimulated eicosanoid release from uninflamed functioning pouch mucosa was not significantly different from that from either ileostomy or defunctioned pouch mucosa. LTB4 and PGE2 release were significantly greater from inflamed than uninflamed pouch mucosa in UC (p = 0.001 and 0.01, respectively). Leukotriene synthesis inhibition or receptor antagonism, or both merit therapeutic evaluation in pouchitis. Increased release of LTB4 from endoscopically normal pouch mucosa suggests increased 5-lipoxygenase activity in patients with UC and could contribute to their predisposition to pouchitis.

Colonic preservation reduces need for parenteral therapy, increases incidence of renal stones, but does not change high prevalence of gall stones in patients with a short bowel

Forty six patients with less than 200 cm of normal jejunum and no functioning colon were compared with 38 patients with similar jejunal lengths in continuity with a functioning colon. Women predominated (67%), and the most common diagnosis in each group was Crohn's disease (33 of 46 no colon, 16 of 38 with colon). All patients without a colon and less than 85 cm of jejunum and all those with a colon and less than 45 cm jejunum needed long term parenteral nutrition. Six months after the last resection 12 of 17 patients with less than 100 cm jejunum and no colon needed intravenous supplements compared with 7 of 21 with a colon. Between 6 months and 2 years, little change occurred in the nutritional/fluid requirements in either group, though there was weight gain. Of 71 patients assessed clinically at a median of 5 years, none with more than 50 cm of jejunum and a colon needed parenteral supplements. Most (25 of 27) of those without a colon who did not need parenteral supplements required oral electrolyte replacement compared with few (4 of 27) with a colon. None of the patients without a colon developed symptomatic renal stones compared with 9 of 38 (24%) with a colon (p < 0.001). Stone analysis in three patients showed calcium oxalate. Gall stone prevalence was high but equal in the two groups--43% of those without and 44% of those with a colon.

Controlled trial of endoscopic injection treatment for bleeding from peptic ulcers with visible vessels

Controlled trials have shown that bleeding peptic ulcers can be successfully treated by endoscopy and injection of adrenaline, with or without sclerosant. However, these trials have been done in major research centres, and endoscopic treatment of upper gastrointestinal bleeding has not yet become routine in general hospitals. We have done a prospective, randomised, controlled trial of injection treatment for bleeding peptic ulcers in a district general hospital. Between April, 1989, and June, 1991, all patients with acute upper gastrointestinal bleeding (n = 555) underwent endoscopy by an experienced endoscopist within 24 h of admission. 98 patients were found to have an ulcer with a visible vessel, of whom 93 were randomised to injection (n = 48) or standard treatment alone (n = 45). Injection treatment consisted of 1-2 ml of 1 in 10,000 adrenaline injected at four to six sites around the ulcer. Adrenaline and 5% ethanolamine oleate (1-2 ml) were then injected directly into the vessel. The medical team managing the patient was unaware of the endoscopic treatment given. The two groups were similar for age, initial haemoglobin concentration, shock, and ulcer site. Rebleeding (injected 8 [16.7%] vs control 21 [46.7%], p = 0.011) was significantly reduced in treated patients. The treated group also had lower mortality (4 [8.3%] vs 9 [20%]), requirement for surgery (4 [8.3%] vs 8 [17.8%]), and mean blood-transfusion requirement (5 vs 7.5 units). Endoscopic injection treatment in our patients significantly reduced rebleeding rate and may have other benefits. This cheap and widely applicable treatment can be used routinely in the management of patients with bleeding peptic ulcers who are at high risk of rebleeding.

Verapamil inhibits in-vitro leucotriene B4 release by rectal mucosa in active ulcerative colitis

Increased mucosal eicosanoid synthesis occurs in active ulcerative colitis; suppression of the synthesis of pro-inflammatory leucotrienes could be therapeutically useful. Neutrophil 5-lipoxygenase is calcium-dependent. In this study, the effect of the calcium channel antagonist, verapamil, on the release of eicosanoids by colitic rectal mucosal biopsies has been examined. Verapamil in therapeutic concentration (5 micrograms/ml, 10(-5) M) reduced leucotriene B4 release from actively inflamed rectal mucosa by 30% (from 60 (5.0 S.E.M.) ng/g wet weight/20 min without, to 42 (5.7 S.E.M.) with verapamil, P less than 0.05), but had no effect on leucotriene B4 release by rectal biopsies taken from patients with quiescent ulcerative colitis (39 (2.8 S.E.M.) ng/g wet weight/20 min without, and 43 (5.0 S.E.M.) with verapamil). Verapamil did not affect mucosal prostaglandin E2 release. The results suggest that, in active ulcerative colitis, verapamil inhibits mucosal 5-lipoxygenase activity and warrants therapeutic evaluation.

Failure of allopurinol to modify urinary composition in enteric hyperoxaluria

Conventional treatment of enteric hyperoxaluria (EHO) consists of dietary restriction of oxalate and fat and correction of its underlying cause whenever possible. Recent work suggests that allopurinol reduces the incidence of urolithiasis and the urinary excretion of both oxalate and uric acid in patients without intestinal disease. We have assessed the effect of allopurinol, 300 mg daily for 2 weeks, on urine biochemistry in patients with EHO due to small bowel Crohn's disease and/or resections. Compliance with treatment was confirmed by a fall in plasma uric acid in every patient. Allopurinol failed to alter 24 h urinary oxalate excretion or oxalate concentration. There were also no significant changes in the urinary excretion of glycollate (like oxalate, a breakdown product of glyoxylate), citrate, magnesium or calcium, each of which was at the lower end of the normal range before and during treatment with allopurinol. It appears unlikely that allopurinol will prove useful in the prevention of urolithiasis in patients with EHO.