Glycaemic responsiveness to long-term insulin plus sulphonylurea therapy as assessed by sulphonylurea withdrawal

To assess the effect of sulphonylurea (SU) in patients with Type 2 diabetes undergoing long-term combination therapy with insulin, by withdrawal of SU, and to identify clinically useful markers of long-term response.

We studied 25 patients, aged 59-83 years, mean glycated haemoglobin (HbA(1c)) 7.0 +/- 0.6%, who had been treated with SU for 16 years (7-24 years) in combination with insulin for 10 years (6-15 years). After basal measurements, SU was withdrawn. Fasting plasma glucose (FPG) and C-peptide were then monitored every 2-3 days during the following 2 weeks. If FPG increased > 40% or P-glucose exceeded 20 mmol/l, SU was restarted. If neither criterion was met, a clinical follow-up visit with measurement of HbA(1c) was scheduled within 8 weeks.

Twenty patients were restarted on SU because of worsening glycaemic control, eight within the first 4 weeks and the remaining 12 at the follow-up visit as their HbA(1c) had increased by 1.1% (range 0.4-2.0%). All these patients were defined as 'SU responders'. The increase in FPG during the initial 2 weeks correlated positively with duration of diabetes (P < 0.01) and duration of SU treatment (P < 0.001). The 'SU responders' had higher levels of basal fasting C-peptide (0.84 +/- 0.44 vs. 0.41 +/- 0.15 nmol/l, P < 0.05), but the variation was wide and none of the measured variables identified 'SU responders'.

In 80% of this group of patients, glycaemic control deteriorated after SU withdrawal despite long duration of SU treatment.