Subcutaneous Continuous Glucose Monitoring: Feasibility of a new microdialysis-based glucose sensor system

Devices for continuous glucose monitoring (1,2) should display the data in real time rather than retrospectively. This is possible by means of the microdialysis technique (3). Therefore, we examined the feasibility of continuous glucose monitoring with functional models of a new microdialysis-based subcutaneous continuous glucose monitoring (SCGM) system (Roche Diagnostics, Mannheim, Germany) for up to 72 h. This system was calibrated by a one-point calibration shortly after the beginning of the recording period.

The SCGM system consists of a microdialysis catheter and a portable extracorporal electrochemical glucose sensor. It displays a new glucose value every minute. The system was tested in 23 ambulatory inpatients with insulin-treated type 1 and type 2 diabetes. For comparison, up to 75 capillary blood glucose samples were taken in each patient. The SCGM signal was corrected for the time that is needed for fluid transportation from the catheter to the extracorporal sensor (31 min) and calibrated by a linear one-point calibration model using the first capillary blood glucose value after an initial equilibration period (mean 4.7 h).

The SCGM system was well tolerated in all patients over the 3-day study. It did not limit the patients’ daily activities. No adverse event occurred, except for a mild skin irritation caused by dressing tape. The sensor signal of the SCGM system showed no system-inborn drift. The percent difference between the SCGM system and the capillary blood glucose was comparable over the observed blood glucose range, with an intra-individual mean absolute difference of 14.8 ± 9.9% (mean ± SD). A representative recording is shown in Fig. 1.

Furthermore, the ability of two glucose spot-measurement schemes to detect hypoglycemic episodes was analyzed in a standardized manner. Glucose spot measurements were retrospectively derived on the basis of continuous glucose profiles according to two different measurement schemes: four times per day (three times preprandially and bedtime) and seven times per day (adding three times 2-h postprandially). Hypoglycemia was arbitrarily defined as glucose <70 mg/dl.

Spot measurement of glucose, performed four or seven times per day, would not detect 71% (22 of 31) or 58% (18 of 31), respectively, of all hypoglycemic episodes registered by continuous glucose monitoring.

Continuous glucose monitoring by this new SCGM system in diabetic patients was feasible and safe for at least 3 days. In contrast to intracorporally placed glucose electrodes (4), the SCGM system showed no time-dependent decline in glucose-sensing sensitivity. Therefore, calibration of the system by just one capillary blood glucose measurement at the beginning of the monitoring period was sufficient.

This study further supports the observation that hypoglycemic episodes in insulin-treated patients are, to a large extent, not detected by spot measurements, even if they are performed seven-times per day (3,5).

In conclusion, the SCGM system offers real-time continuous glucose monitoring with sufficient precision in diabetic patients over 72 h. Detecting asymptomatic hypoglycemic episodes by continuous glucose monitoring will facilitate the management of patients with diabetes.