A previously undescribed side effect of icodextrin: overestimation of glycemia by glucose analyzer

Serious discrepancies between glycemia measurements obtained with an Accutrend Sensor (Boehringer Mannheim GmbH, Mannheim, Germany) type analyzer (based on a glucose dehydrogenase enzymatic reaction) and measurements obtained in the laboratory by a reference method (hexokinase) have been found in an insulin-requiring, diabetic, continuous ambulatory peritoneal dialysis (CAPD) patient treated with icodextrin 7.5% (Extraneal; Baxter Healthcare SA, Castlebar, Ireland), a new osmotic agent for peritoneal dialysis. We therefore investigated the respective role of the Analyzer and of the glucose polymer in this hitherto undescribed problem.

Glycemia was measured simultaneously on venous blood using a reference laboratory technique, and on capillary blood using the Accutrend Sensor glucose analyzer in three groups of CAPD patients: 6 patients on Extraneal for at least 1 week, 6 patients receiving their first Extraneal exchange, and 8 patients never exposed to Extraneal. In the first group of patients, glycemia was also measured with another analyzer (Glucocard; Menarini Diagnostics, Firenze, Italy) using a different enzymatic reaction (glucose oxidase). In a separate study, whole blood of a normal subject was spiked with concentrated solutions of glucose and icodextrin and some of its metabolites (maltose, maltotriose, maltopentaose). Once again, comparative measurements of glycemia were performed with the Accutrend Sensor, with two other kits using a glucose dehydrogenase enzyme reaction, and with the hexokinase reference method.

In 6 CAPD patients treated with once-daily exchanges with Extraneal for a minimum of 7 consecutive days, we confirmed overestimation of glycemia by the Accutrend Sensor of 65 +/- 26 mg/dL compared to reference values (p < 0.01), and of 69 +/- 25 mg/dL (p < 0.001) compared to measurements obtained with the Glucocard monitor. In 6 other CAPD patients studied at the end of one single icodextrin exchange, overestimation of 61 +/- 11 mg/dL was already present (p < 0.001). On the other hand, in 8 CAPD patients never treated with icodextrin, there was no discrepancy between the Accutrend Sensor readings and reference values. The measurements in spiked blood confirmed that only the Accutrend Sensor overestimates glycemia in the presence of maltose and glucose polymers. The overestimation decreased as the molecular size of the saccharides added to blood increased. There was no overestimation when other kits using a dehydrogenase enzyme were tested.

The overestimation observed is probably related to the presence of oligosaccharides (mainly maltose), derivatives of glucose polymers present in Extraneal and absorbed via the peritoneal route, in the blood of patients treated with icodextrin. The glucose dehydrogenase characterizing the Accutrend Sensor, an enzyme of the pyrroloquinolinequinone class, very likely reacts with the free reducing group of the glucose molecule located at the end of each saccharide chain. This would not be the case for the Glucocard monitor using glucose oxidase, for other kits using glucose dehydrogenase, and for the reference method based on hexokinase. The Accutrend Sensor type of analyzers are therefore not suitable for regular monitoring of glycemia in diabetic PD patients treated with icodextrin.