Aims Induction regimes for ALL incorporate high dose steroids and L-asparaginase; both known to affect glucose homeostasis. Previous studies report hyperglycaemia in >10% of patients during induction, but the impact of hyperglycaemia on treatment outcome remains uncertain. We reviewed the incidence of hyperglycaemia in a retrospective cohort of children with ALL and correlated the findings with the risk of infection and death during induction.

Methods We reviewed records of all patients diagnosed with ALL at a paediatric oncology Principal Treatment Centre between December 2003 and May 2011. 144 patients met the criteria: under 16 years at diagnosis; no pre-existing diabetes; treatment on a single national protocol (UKALL 2003). Blood glucose records obtained during the first 28 days of therapy were reviewed. Patients were categorised according to the highest glucose level measured. Mild hyperglycaemia was defined as blood glucose 7.7-11.1 mmol on two or more days, and overt hyperglycaemia as blood glucose >11.1 mmol on more than one day. All other patients were euglycaemic. Case notes, pathology systems and clinical trial records were used to identify suspected (new febrile illness) and proven (culture positive) infections and survival.

Results 36% (52) patients had overt and 33% (48) mild hyperglycaemia. Seven (5%) of all patients had Down's syndrome. The risks of infection and induction death across the three groups are shown below (table 1). Subset analyses are shown for Down's syndrome patients.

Infection was more common in the overt hyperglycaemia group: proven infection occurred in 16/52 (31%) in this group compared with 5/44 (11%) of the euglycaemia group (RR 2.7; 95% CI 1.1-6.8, p=0.04). There were two induction deaths, both from sepsis and both in patients with overt hyperglycaemia. Overt hyperglycaemia was very common in Down's syndrome (6/7) patients, who had the highest incidence of proven sepsis.

Conclusion This study confirms hyperglycaemia as an important complication of induction treatment for ALL and that patients with overt hyperglycaemia are more likely to develop infection. Further analyses are ongoing to identify risk factors for hyperglycaemia. Future studies are needed to assess the benefit of stricter glycaemic control. Patients with Down's syndrome are at highest risk.