Profil Blog

Downtime Impact Factor evaluates downtime in automated glucose clamps

Written by Mareike Kuhlenkötter | Nov 14, 2024 2:45:00 PM

The euglycemic hyperinsulinemic glucose clamp technique is the gold standard for the determination of pharmacokinetic (PK) and pharmacodynamic (PD) effects of (new) glucose lowering drugs, insulins in particular. The measurement of a subject’s BG concentration and the resulting regulation of BG by means of varying glucose infusion rates (GIR) can either be conducted manually or automatically. According to a guideline of the European Medicines Agency [1] both methods have been reported to provide similar and reproducible results as long as there are no rapid changes in glucose requirements. 

All PD parameters used to describe the effect of a glucose lowering substance, in particular total area under the curve (AUC), maximum GIR (GIRmax) and time to maximum GIR (tGIRmax), are determined on the basis of the GIR-curve. 

However, most of the clamp quality parameters [2], provided to estimate the performance of the clamp, are BG based.(Link zu Blog Glucose Clamp Quality?) The sole exception is the quality parameter Utility, describing the percentage of operational time.

Operational downtime means that the regular mode of measuring BG every minute and calculating and infusing the right amount of glucose is not operating, for example due to urination during the experiments. In clinical practice we adjust the GIR manually, if needed.

We investigated the impact of downtime on the PD parameters GIRmax and tGIRmax, which are particularly relevant in studies analyzing biosimilars.

Methods

We used a single compartment model to perform numerical simulations. The blood glucose lowering effect was induced by an insulin profile obtained from the real clamp data of a subject in a clinical trial.

Technical downtimes of varied length were simulated by interrupting the automated adaptation of the GIR. However, BG changes were generated every minute as the simulated BG-lowering effect of the different simulated insulins continued.

Although the clamp quality parameters (precision (CV), control deviation (CD), and (utility) performed well in the simulated clamps, the PD parameters (GIRmax, tGIRmax) were disturbed.

Right after the end of the simulated downtimes in operation, when the BG concentration was not automatically controlled and managed, greater differences between BG concentration and target level occurred. Larger differences have more impact on the PD parameters. Obviously, the longer the downtime lasts, the greater the risk for infusing a less than optimal glucose amount and, therefore, disturbing the PD parameter.

This effect can be detected in BG measurements after the downtime. We therefore introduced the Downtime Impact Factor (DIF), describing the difference between ‘utility’ and ‘resulting utility’, as results are only considered if they are in a 10 percent range around the target level. 


Results

By means of in-vitro experiments we validated the results of the numerical simulations. We tested the same insulin profile several times solely with manually induced downtimes at identical times, but different infusions of glucose. In the figure below measured GIR-curves are displayed, which lead to different PD parameters. For comparison reasons an undisturbed GIR-curve is shown in black (clamp1).

The blue GIR-curve (clamp2) resulted when the operator adapted the infused GIR effectively, whereas the light blue curve (clamp3) resulted if the operator overreacted slightly. The green curve (clamp4) was measured if the operator did not react. Although the quality parameter were acceptable in every clamp, tGIRmax differed. Therefore, DIF greater than 5% is a robust marker for the detection of disturbed GIR-curves, considering deviations of less than 10% between disturbed tGIRmax and real tGIRmax as acceptable.

DIF was applied to all clamps that occurred in an automated glucose clamp study with 26 subjects participating in two clamps each (37492 minutes of automated clamp data). Overall, 367 technical downtimes were identified with a total downtime of 1965 minutes with downtime periods between 1 and 32 minutes. None of the clamps had a DIF greater than 5%. As clinic staff is trained to check BG concentration during downtimes, manual corrections of GIR during downtimes were performed optimally.


Conclusion

Clamp quality parameters are established to provide information about the performance of clamps on the basis of BG results. However, PD parameters describe characteristics of the resulting GIR curve. DIF is the first quality parameter that evaluates the effect of BG excursions from the GIR curve. DIF greater than 5% is a robust marker that PD parameter might be affected by too large differences between BG concentration and BG target level.