SUMMARY The effects of cytotoxic therapy on the structure and function of the proximal jejunum were studied in six patients receiving intravenous cyclophosphamide (300 mg/M2), methotrexate (40 mg/M2), and 5-fluorouracil (600 mg/M2) as adjuvant therapy for breast cancer. Using a steady state, triple lumen tube perfusion system the absorption of water and electrolytes was measured before and 48 h after administration of the cytotoxic agents. Jejunal biopsies were obtained at each perfusion. Median (range) water absorption fell from 126 (40-142) to 84 (46-142) ml/h/30 cm, with parallel changes for electrolytes; none of the changes was significant. Brush border disaccharidases did not change at 48 h after chemotherapy, while mature enterocytes appeared normal by both light and electron microscopy. Crypt cells and immature enterocytes, however, showed focal vacuolation by light microscopy, corresponding to the occurrence of large residual bodies (secondary lysosomes) containing partially degraded fragments of damaged crypt cells. The confinement of ultrastructural changes to the immature cell population may explain the failure of this study to show a consistent change in the absorptive function of the jejunum 48 h after chemotherapy.Cytotoxic therapy can cause structural and functional changes in the human proximal intestine. '-3 This effect may be important in the development of gastrointestinal symptoms such as diarrhoea and may lead to malabsorption of orally administered drugs, including cytotoxic agents.4 5 Smith et a!6 could find no change in xylose absorption and faecal fat estimation in patients with biopsy proved depression of crypt cell mitosis, but most previous studies of cytotoxic damage have examined changes of gastrointestinal function and structure independently.The technique of small intestinal perfusion is a sensitive method for detecting changes in the transport of water and electrolytes across the jejunal mucosa and might therefore show relatively minor abnormalities of enterocyte function more readily than conventional intestinal absorption tests.7 We have used small intestinal perfusion to investigate IPresent address:

SUMMARY Fifty two patients with Crohn's disease (31 outpatients and 21 inpatients) were investigated for evidence of vitamin A deficiency. Eleven (21%) had low plasma retinol concentrations (<1.2 umol/l (34.3 ,ug%)). Five of these were outpatients and plasma retinol was only slightly reduced (>1.0 ,umoWl (28.6%)). All outpatients weighed 80% or-more of ideal, and were considered at low risk of developing vitamin A deficiency. In contrast, of the six inpatients with low plasma retinol concentration, five had a level of <1*0 ,umoUl (28.6 ,ug%) and weighed <80% ideal. Three of these had impaired dark adaptation and a plasma retinol concentration of <0.8 ,umol/1 (<22.9 ,ug%). As a group, the inpatients were more protein depleted than the outpatients, with respect to serum albumin (p<0-01), transferrin (p<0-001), and prealbumin (p<0.001) but retinol binding protein levels were not significantly lower. It is suggested that patients with extensive small bowel Crohn's disease, who weigh <80% of ideal weight, merit measurement of plasma retinol concentration. Those with plasma retinol <0.8 ,umol/l (<22.9

We calculated respiratory system resistance (Rrs) and elastance (Ers) from pressure and flow at the mouth in six seated subjects relaxed at FRC (cheeks tightly compressed) during sinusoidal volume forcing (250, 500, and 750 ml) at 0.2, 0.4, and 0.6 Hz. Dependencies of Rrs and Ers on frequency and tidal volume were generally the same in each subject; Rrs tended to decrease with frequency and tidal volume, whereas Ers tended to increase with frequency and decrease with tidal volume. Multiple linear regression of combined data indicated that the frequency and tidal volume effects on Rrs and Ers were significant (p less than 0.05), and that the effects on Rrs decreased at higher flows. Average Rrs was highest (4.43 cm H2O/L/s +/- 0.21 SE) at 0.2 Hz-250 ml, and lowest (3.07 cm H2O/L/s +/- 0.37) at 0.6 Hz-750 ml. Average Ers was highest (12.1 cm H2O/L +/- 1.1) at 0.6 Hz-250 ml, and lowest (7.1 cm H2O/L +/- 0.6) at 0.2 Hz-750 ml. We conclude that frequency and tidal volume dependencies in Rrs and Ers in the normal range of breathing should be considered when interpreting measurements of respiratory system impedance or developing models to describe the mechanical behavior of the respiratory system.

SUMMARY Serum zinc concentrations and urine zinc excretion have been studied in 10 patients with severe Crohn's disease before and during 59 patient-weeks of intravenous nutrition. Before serum zinc concentrations (9.9±1.0 ,umol/l: mean ± SEM) and urine zinc excretion (3.3±0±6 gmol/24h) were less than controls (p<001). No patients had clinical signs of zinc deficiency before intravenous nutrition and none developed signs during it. There was no overall change in serum zinc concentrations, despite improvements in body weight, skinfold thickness, and mid-arm circumference in all patients, and increased serum albumin and serum transferrin concentrations during all but two periods of intravenous nutrition. Nor was there any relationship between serum zinc concentrations and zinc intake (up to 220 ,umoUday), serum zinc concentrations remaining significantly lower than control levels. Urine zinc excretion during the first week of intravenous nutrition showed a 1.2 to 53-fold increase (mean 11-fold) over pre-intravenous nutrition levels, and a positive relationship was demonstrated between zinc intake and urine zinc excretion. It is suggested that zinc supplied by the intravenous route is inefficiently transported to the tissues, and that some is excreted in the form of small molecular weight chelates into urine. Recommendations are made for the supply of intravenous zinc, based on monitoring urine zinc excretion in individual patients.

No abstract