Hyperglycaemia complicating haemolytic uraemic syndrome

J. Paediatr. Child Health (1990) 26, 225-226

Hyperglycaemia complicating haemolytic uraemic syndrome B. A. CRAWFORD, L. P. ROY and J. F. KNIGHT The Children’s Hospital, Camperdown, New South Wales, Australia

Abstract The occurrence of hyperglycaemia and insulin deficiency in a young child receiving peritoneal dialysis during the course of haemolytic uraemic syndrome (HUS) is described. This unusual complication may have been due to microvascular disease involving the pancreas. Plasma glucose should be monitored during HUS, particularly if dialysis with fluids containing high dextrose concentrations is required.

Key words:

haemolytic uraemic syndrome; hyperglycaemia; insulin deficiency; peritoneal dialysis.

Although the kidney is the organ most commonly affected by the small vessel thrombosis characteristic of haemolytic uraemic syndrome (HUS), other organs may be involved. In one extensive autopsy series, the pancreas was the second most common extrarenal organ involved (20 of 26 cases) after the colon.’ Nevertheless, clinical evidence of hyperglycaemia and insulin deficiency in patients with HUS is rare. From a series of 80 cases of HUS seen at The Children’s Hospital, Camperdown between 1966 and 1989, the patient reported here is the only one recognized as demonstrating this complication.

CASE REPORT

A girl aged 2 years 10 months presented to a country hospital with a 1-day history of bloody diarrhoea. Her past history was unremarkable apart from alopecia totalis of 15 months duration. Abdominal pain and vomiting developed so a laparotomy for suspected bowel obstruction was performed. This revealed an inflamed oedematous sigmoid colon and a sigmoid volvulus which was untwisted. The following day she became anuric and oedematous. Laboratory investigations showed creatinine 317 pmollL, haemoglobin 7.3 g/dL, platelets 50 OOO/mm3 and fragmented red blood cells were observed in the peripheral blood film. The diagnosis of HUS was made and she was transferred to The Children’s Hospital, Camperdown. On admission, physical examination revealed an alert child, weight 15.5 kg, blood pressure 125/85, generalized peripheral oedema and a distended abdomen with ascites. Peritoneal dialysis was commenced using hourly cycles of 500 mL of 4.25% dextrose dialysis fluid. A Dextrostix reading taken just prior to the commencement of dialysis was 12.6 mmol1L. The patient’s intake was nil orally and intravenous fluids (N14 saline + 3.75%

~

~

Correspondence Dr J F Knight, The Children’s Hospital. Pyrmont Bridge Road, Camperdown, NSW 2050, Australia 6 A Crawford, Paediatric Registrar, MB, BS, L P Roy, Head, Department of Nephrology, MB, BS, BSc(Med), FRACP. and J F Knight, Staff Physician in Nephrology. BA, MB, BS, FRACP Accepted for publication 1 June 1990

dextrose) to replace insensible losses only. Twenty-six hours after the initiation of dialysis the Dextrostix reading was 10.9 mmolIL and glucose monitoring was ceased. The highest level recorded during this period was a plasma glucose of 20.3 mmol1L. Insulin was not given. Dialysis continued using 2.5 or 4.25% dextrose dialysis fluid with dwell times of 3 h or less. On the 10th hospital day, owing to persisting problems with fluid overload and hypertension, the dialysis programme was changed to 30-min cycles using 4.25% dextrose dialysis fluid. Twelve hours later, having had no oral intake for 10 h, the patient became acutely confused and disorientated with writhing movements and facial grimaces of such force as to cause tongue lacerations from biting. This was associated with brief episodes of bradycardia. She had not received any drugs known to cause dystonic reactions. On examination there were no focal neurological signs but generalized hyperreflexiawith bilateral ankle clonus was present. Investigations revealed sodium 130, potassium 4.9, chloride 90, bicarbonate 11, urea 30 mmollL, creatinine 405 pmollL, plasma glucose 82.2 mmollL, calcium 2.1 1 mmol/L, albumin 21 glL, serum amylase 118 UIL (normal 15-1 10 UIL). A plasma insulin level obtained at this time was 11 mU1L. Ketoacidosis was indicated by the results of an arterial blood sample: pH 7.34, Pco, 28.1 mmHg, Po, 85.7 mmHg, base excess -9.7 mm1L and a calculated anion gap of 34. An electroencephalogram (EEG) was consistent with a toxic metabolic encephalopathy . The dialysis fluid dextrose concentration was changed to 1.5%, and 5 U of short-acting insulin was added to each 500 mL of dialysate. An intravenous insulin infusion (0.05 UIkg per h) was administered until the hyperglycaemia and acidosis were corrected and the level of consciousness had returned to normal. For the next 4 weeks blood glucose levels ranged from 2.2 to 20.9 mmol1L requiring the addition of up to 2 U actrapid insulin to each bag of dialysis fluid. Supplemental insulin was not required after the 36th hospital day. There was no family history of diabetes mellitus. Prior to discharge an oral glucose tolerance test (GTT) (1.75 glkg glucose load) was performed following an overnight fast. The fasting plasma glucose was 4.6 rnmollL and there was a normal glycaemic response following oral glucose challenge as determined by criteria established by the National Diabetes Data Group.‘ Insulin levels rose appropriately during the GTT. A renal biopsy performed on day 23 showed 50% of glomeruli

226

had haemorrhagic necrosis, with focal basement membrane thickening in the remaining glomeruli. lmmunofluorescence did not reveal any evidence of immunoglobulin or C3 deposition. Serum antinuclear antibody was negative and complement levels were normal. Stool cultures failed to grow pathogens and assay for verocytotoxin was n e g a t i ~ e . ~ Captopril was commenced for the management of hypertension. Due to the development of a stricture in the sigmoid colon a defunctioning transverse loop colostomy was performed which was repaired at a subsequent admission. Dialysis was ceased 36 days after admission. Twelve months later she remains well on captopril with a glomerular filtration rate (measured by DTPA clearance) of 47 mLlmin per 1.73 m2.

B. A. Crawfordet a/.

also observed and ketoacidosis is rare. Pancreatitis has recently been reported to be associated with hyperglycaemia in high dextrose peritoneal dialysis.'2 This patient's serum amylase level, however, was not indicative of pancreatitis. The amylase level was not measured in the dialysis fluid. Alopecia totalis is part of the autoimmune polyglandular syndrome type II, which also includes type I diabetes mellitus. Moreover, a recent report suggests that HUS may be mediated by antibodies to endothelial cells.13The routine immunological testing this child underwent does not exclude the possibility of a multi-organ auto-immune diathesis. Hyperglycaemia due to acute insulin deficiency may occur in patients with HUS, presenting a sudden and life-threatening complication. Blood glucose levels should be monitored during the acute phase of HUS, particularly if dialysis with fluids containing a high dextrose concentration is required.

DISCUSSION

This child developed severe but transient hyperglycaemia associated with an inappropriately low insulin level during the course of HUS. The absence of this complication in the previous extensive experience of HUS at this hospital suggests that symptomatic hyperglycaemia is an unusual complication of HUS. A review of the literature reveals only six previous case reports4-' However, in one centre over a 4-year period, three of 21 patients with HUS developed hyperglycaemia requiring insulin the rap^.^ In all reported patients, hyperglycaemia occurred during the acute phase of the illness. Three of the six reported cases documented low insulin levels in association with hypergly~aemias.~ In the others, the combination of hyperglycaemia and ketoacidosis was consistent with a lack of insulin. The clinical course of these six reported patients suggests that hyperglycaemia is a poor prognostic indicator in HUS. Three patients died and two remained insulin dependent diabetics, one of whom also had residual renal impairment requiring continuing dialysis. The majority of histological reports from patients with HUS describe vascular changes in the pancreas restricted to the islets.',6x8This may be because the islets of Langerhans are more energy dependent than the exocrine portion of the gland. However, necrosis of the entire pancreas has been reported.' The high concentration of dextrose in the dialysis fluid used in this case may have accentuated the hyperglycaemia although mild hyperglycaemia was noted even before commencing dialysis. Dextrose dialysate at 4.25010 has a dextrose concentration of approximately200 mmollL, providing a positive concentration gradient across the peritoneumeven when plasma glucose levels are high. Although hyperglycaemia is a recognized complication of peritoneal dialysis it is usually mild and of no clinical significance, although hyperosmolar hyperglycaemic nonketotic coma has been r e p ~ r t e dIn . ~this circumstance elevated insulin levels might be expected. Uraemia and stress are causes of hyperglycaemia.'O,l' In these circumstances, however, high serum insulin levels are

ACKNOWLEDGEMENTS

The authors wish to thank Dr P. Doyle, Port Macquarie, for referral of this patient, Dr C. T. Cowell for endocrinological consultation, Dr M. Gapes for performing the verocytotoxin assay and Miss Michelle Giaquinto for preparation of the manuscript.

REFERENCES 1 Gianantonio C. A,, Vitacco M.. Mendilaharzu F., Gallo G. E., Sojo E. T. The hemolytic uremic syndrome. Nephron 1973, 11: 174-92. 2 National Diabetes Data Group Classification and diagnosis of diabetes mellitus

and other categories of glucose intolerance Diabetes 1979; 28: 1039-57 3 Konowalchuk J., Speirs J. I , Stavric S. Vero response to a cytotoxin of Escherichla coli. Infect. Immun. 1977; 18. 775-9. 4 Adelman R. D , Halsted C. C.. Sheikholishlam 6.M. Hemolytic uremic syndrome' Associated conditions. J Pediatr. 1980; 97. 161 5 Andreoli S. P I Bergstein J. M.. Developmentof insulin dependent diabetes mellitus during the hemolytic uremic syndrome J Pediatr. 1982, 100' 541-5 6 Burns J C.. Berman E. R.. Fagre J L., Shikes R. H., Lum G. M. Pancreatic islet cell necrosis: Association with hemolytic uremic syndrome J. Pediatr. 1982; 100. 582-4. 7 Primhak R. A , Taitz L. S., Variend S.,Webb D. H. K., Cser A. Necrosis of the pancreas in the haemolytic uraemic syndrome. J. Clin. Pathoi. 1984; 37:655-8 8 Shumway C. N , Terplan K. L. Hemolytic anemia, thrombocytopenia and renal disease in childhood. Pediatr. Clin. N. Amer. 1964; 11: 577-90. 9 Boyer J. M Gill G. N.. Epstein F. H. Hyperglycaemia and hyperosrnolality complicating peritoneal dialysis. Ann. Intern Med. 1967; 67: 568-71. 10 DeFronzo R. A., Smith D., Alvestrand A. Insulin action in uremia. Kidney Int. 1983. 24. S102-14 11 Hamburg S.,Hendler R., Sherwin A. S. Influence of small increments of epinephrine on glucose tolerance in normal humans. Ann. Intern. Med. 1980; 93: 566-8. 12 Emder P. J , Howard N J., Rosenberg A. R . Non-ketotic hyperosomolar diabetic pre-coma due to pancreatitis in a boy on continuous ambulatory pentoneal dialysis. Nephron 1987; 44: 355-7. 13 Leung D. Y. M., Moake J. L.. Havens P. L., Kim M., Pober J. S. Lytic antiendothelial cell antibodies in haemolytic uraemic syndrome. Lancet 1988; ii: 183-6.

.