I.W. Booth

Hazards of parenteral treatment: do particles count?

After prolonged parenteral nutrition a 12 month old infant died with pulmonary hypertension and granulomatous pulmonary arteritis. A review of necropsy findings in 41 infants who had been fed parenterally showed that two of these also had pulmonary artery granulomata, while none of 32 control patients who died from sudden infant death syndrome had similar findings. Particulate contaminants have been implicated in the pathogenesis of such lesions and these were quantified in amino acid/dextrose solutions and fat emulsions using automated particle counting and optical microscope counting respectively. Parenteral feed infusions compounded for a 3000 g infant according to standard nutritional regimens were found to include approximately 37,000 particles between 2 and 100 microns in size in one days feed, of which 80% were derived from the fat emulsion. In-line end filtration of intravenous infusions may reduce the risk of particle associated complications. A suitable particle filter is required for use with lipid.

Faecal chymotrypsin: a reliable index of exocrine pancreatic function.

Simultaneous measurements of duodenal and faecal chymotrypsin were made in 30 children aged 3 weeks to 14 years. Apparent chymotrypsin secretion rates measured after stimulation with pancreozymin were compared with the mean faecal chymotrypsin concentration derived from three stool specimens collected at random within 72 hours of the intraduodenal test. In the 25 children who responded to pancreozymin stimulation the mean faecal chymotrypsin concentration was significantly positively correlated with the apparent chymotrypsin secretion rate. Correlation using single specimen stools collected at random was appreciably poorer. In the five children with undetectable or only traces of chymotrypsin in the duodenum after stimulation, the mean faecal chymotrypsin concentrations were only 3-10% of the lower limit of the reference interval. In a second group of 46 children with cystic fibrosis proved by sweat tests and clinical evidence of malabsorption, the chymotrypsin concentration measured in a single stool specimen collected at random was unequivocally subnormal in each case. Faecal chymotrypsin measurement is a rapid, simple, cheap, readily repeated, non-invasive test of high specificity and sensitivity. Faecal chymotrypsin should be measured before contemplating intraduodenal tests of pancreatic function.

Egg and breast milk based nitrogen sources compared.

A nitrogen source based on egg protein (Vamin 9 glucose) and an alternative with an amino acid profile more similar to breast milk (Vaminolact), were compared in 14 parenterally fed infants. Subjects were randomly allocated to receive one or other amino acid solution, but were otherwise given identical diets. At the start of the study the two groups did not differ significantly in postconceptual age, postnatal age, or weight. Over a six day study period on a stable intake of intravenous nutrients there was no significant difference in growth or nitrogen retention between the two groups. Plasma amino acid profiles in those receiving Vamin 9 glucose, however, were frequently abnormal. Notably, mean concentrations of potentially neurotoxic phenylalanine and tyrosine were significantly higher (140% and 420%, respectively) in patients fed Vamin 9 compared with those given Vaminolact. An amino acid solution based on the composition of breast milk protein therefore brings plasma amino acid profiles during parenteral nutrition closer to those found in breast fed infants, and reduces in particular, the risks of hyperphenylalaninaemia and hypertyrosinaemia.

Simplified oral pancreatic function test.

The standard Bentiromide test and a new modified test using p-aminosalicylic acid (PAS) as a pharmacokinetic marker for p-aminobenzoic acid (PABA) have been evaluated in the detection of pancreatic exocrine insufficiency in children. The conventional two day test using a colorimetric assay for urinary PABA discriminated poorly between five children with pancreatic insufficiency and 13 others with normal pancreatic function. Two further groups of patients, comprising 28 with pancreatic exocrine insufficiency and 20 with normal pancreatic function underwent the modified test, and urine samples were analysed by high performance liquid chromatography. The results showed a complete separation between groups. The use of PAS eliminates a number of sources of error inherent in a two day Bentiromide test and provides a simplified and accurate diagnostic test for pancreatic insufficiency. The PABA-PAS modified test enables collection of the urine to be done during a single six hour period.

Using the steatocrit to determine optimal fat content in modular feeds.

The steatocrit was measured in infants with protracted diarrhoea who were receiving intragastric modular feeds. Measurements were made when fat intake was constant to determine steatocrit variability and during increases in fat intake to determine fat tolerance limits. Steatocrit variability was expressed as the range between a subjects lowest and highest steatocrit value. The median between-stool and between-day variabilities were 2% and 11% respectively. Variability was also measured in seven healthy breast fed infants. In the five still displaying appreciable physiological steatorrhoea, the between-day variability (median 8%) and between-stool variability (median 9%) were not significantly different from the protracted diarrhoea group. In the fat tolerance investigations in the protracted diarrhoea group, the steatocrit increased with increases in the module fat content. Fat intake and steatocrit were significantly positively related. A significant negative correlation was seen between steatocrit and weight gain, the latter becoming negligible at steatocrit values around 30%.

Risk factors for high diarrhoea frequency: a study in rural Zimbabwe

Diarrhoea morbidity data were collected prospectively over 22 months from a cohort of young children living in a deprived community in rural Zimbabwe. Despite the general high prevalence of diarrhoeal disease, there was considerable individual variability in attack rates. Risk factors associated with high diarrhoea frequency were therefore sought by a questionnaire study on feeding, environmental, educational and socio-economic factors. This was supported by observation of living conditions, and water and sanitation facilities. Surprisingly, no association was found between diarrhoeal morbidity and any of these factors, suggesting that other factors such as individual hygiene behaviour or individual susceptibility to diarrhoea may play a role in determining the observed differences in diarrhoea rates in this community.

PAEDIATRIC GASTROENTEROLOGY: Dietary management of acute diarrhoea in childhood

Gastroenteritis in children is usually treated with the graded introduction of milk feeds after rehydration. Although having never been rigorously tested the practice of gradually increasing milk strength over several days has been considered an appropriate means of warding against lactose intolerance and preventing sensitization to cows milk antigens. These guidelines were formulated in Europe and North America and invariably lead to a reduction in nutrient intake. Malnourished children in developing countries however may experience an average 5-6 episodes of acute diarrhea per year and the nutrient effects are cumulative. A recent study from Latin America explored whether continued feeding is safe for infants under age 6 months and whether malnourished children respond adversely. Infants randomly assigned to receive full strength cows milk immediately after rehydration did not have more treatment failures higher stool outputs or longer lasting diarrhea than those whose feeds were regarded to full strength over 48 hours. It is unclear however whether the youngest or more malnourished subjects were overrepresented in the treatment failures. Results also indicate that deciding to change treatment should not be dictated by the presence of reducing substances in the faeces; the majority of infants with reducing substances in their stools did well. This study offers the first scientific support for rapidly reintroducing full-strength milk formula after gastroenteritis is malnourished patients under 6 months of age. The 10% of infants in which dehydration recurs after reintroducing milk feeds are still difficult to manage. In the absence of yogurt or lactose-free formula a locally-produced modular feed of chicken starch and vegetable oil may be suitable.

17 KINETICS OF HUMAN SMALL INTESTINAL BRUSH BORDER MEMBRANE D-GLUCOSE TRANSPORT: JEUNUM AND ILEUM COMPARED

Translocation of glucose across the small intestinal brush border membrane is of fundamental physiological and clinical importance, yet little is known about the kinetics of this transport process in man. Using a recently developed miniaturised technique employing human brush border membrane vesicle (BBMV), we have studied the initial 3 sec uptake of glucose by BBMV under Na+ gradient (out:100mM; in:0) and non-Na+ gradient (out and in:0) conditions, over a range of substrate concentrations (extra-vesicular glucose:0-3.2mM). BBMV were derived from small (150-175 mg), full thickness specimens of jejunum (n=8) or ileum (n=12), obtained from children undergoing elective intestinal resection. Eadie-Hofstee transformation of the data showed that in the jejunum, glucose was transported by 3 different mechanisms:(i) a diffusional process and/or a low affinity, Na+-independant carrier(ii) a high affinity, low capacity, Na+-dependant saturable system (approximately Km 0.3mM; Vtaax 13nmol/mg/min).(iii) a Na+-dependant diffusional process and/or a very low affinity, high capacity saturable carrier.In contrast, it was not possible to consistently demonstrate saturable glucose uptake by ileal BBMV. These data provide a basis for an explanation of the heterogeneity seen in glucose-galactose malabsorption, and suggest that in the ileum, transcellular glucose transport may be relatively unimportant.

86. REDUCTION IN NEONATAL HYFHWENYLAIMMAFKIA WITH A LOW PHENYL-ALANINE AMINO ACTD SOURCE

We have recently drawn attention to the worrying incidence of hyperphenylalaninaemia in parenterally fed neonates (1), and have therefore compared Vamin 9 Glucose (KabiVitrum; 5.5 g/1 phenylalanine) with Vaminolac (KabiVitrum; 2.7 g/1 phenylalanine), a modified amino acid solution based upon the composition of human breast milk. Neonatal surgical patients requiring total parenteral nutrition were randomly allocated to receive either Vamin or Vaminolac as the amino acid source. 13 patients (A gastroschisis, 4 small bowel atresia, 3 necrotising enterocolitis, 2 tracheo-oesophageal fistula) were given Vamin and 6 (4 NEC, 1 gastroschisis, 1 jejunal volvulus) received Vaminolac. Amino acid intake was stabilised at 2.5g/kg/ day for a minimum of 3 days prior to plasma amino acid analysis, performed twice weekly for one week, then at weekly intervals. The median plasma phenylalanine concentration in the Vamin group was 200Umol/l (range 49-983umol/l, n=32) compared with only 68umol/l (range 47-154umol/l, n=13; p < 0.00003) even though patients receiving Vamin were of significantly greater post-conceptual age at the time of sampling (median 38 weeks, range 31-53 weeks) than Vaminolac patients (median 38 weeks, range 31-41 weeks; p< 0.00003). No patients had grossly elevated plasma tyrosine concentrations or evidence of liver dysfunction. Three of the neonates given Vamin were found to have plasma phenylalanine concentrations above the commonly accepted neuro-toxic level of 600umol/l. The use of Vaminolac compared with Vamin 9 Glucose is therefore associated with a significant reduction in potentially toxic plasma phenylalanine concentrations.(1) Puntis JWL et al., Lancet 2:1105-1106 (1986).