Model methods and limitations
What AyFlux calculates, how to read the estimates, and where the model has limits.
How the model works
AyFlux uses a two-compartment model to describe how acyclovir distributes and is cleared over time. The central and peripheral compartments are mathematical simplifications, not direct maps of specific organs.
Changing a parameter, such as assumed kidney clearance, changes the predicted curve. The result depends on the model and its inputs, not on measurements from a particular patient.
Plasma and CSF estimates
Concentrations in blood and cerebrospinal fluid (CSF) do not necessarily rise and fall at the same rate. AyFlux displays separate plasma and CSF estimates to explore that difference. The CSF curve is a prediction, not a laboratory measurement.
What is CMMG?
9-Carboxymethoxymethylguanine (CMMG) is a metabolite of acyclovir. Studies have found an association between CMMG concentrations and neurological symptoms in some patients receiving acyclovir.
That association does not make a single concentration a reliable diagnostic cutoff. AyFlux's CMMG curves and reference lines should not be used to diagnose or rule out neurotoxicity. See the CMMG study by Helldén and colleagues and research on acyclovir and CMMG in the central nervous system.
Uncertainty and sensitivity
Shaded bands show how predictions vary under the assumptions used in the simulation. They do not guarantee that an individual patient's concentration will fall inside the band.
Sensitivity analysis shows which inputs have the greatest effect on a prediction. It helps you understand the model, but does not confirm that its assumptions are correct.
What the model cannot tell you
- It cannot measure a patient's actual plasma, CSF, or CMMG concentrations.
- It cannot determine whether an individual has acyclovir neurotoxicity or adequate antiviral exposure.
- It cannot account for every difference in physiology, drug handling, or dialysis performance.
- It cannot establish a safe dose or treatment schedule.
Sources and code
Open the simulator and select Methods and references for the source list and model details. The AyFlux source code is also available on GitHub.