a question related to aucint.last after multiple dosing

[JoannaPeng]

I have a question related to aucint.last after multiple dosing. I have a subject whose conc vs time profile at steady state is shown below. This subject's first two samples at steady state were missing, and thus AUClast and other related parameters were not calculated (please see the 2nd screenshot; which is understandable). The last sample (after Day 35) shown in the plot below was BLQ, so Tlast=14 days. Tmax=3 days and thus, half.life could not be calculated because of too few points (understandable). What I don't understand is why aucint.last with the interval of 0-21 days was estimated (the last row). I don't think it is estimable in this case. Any explanation?

Thank you very much,
Joanna

[JoannaPeng]

Another related question. If the pre-dose sample is missing after a single dose or after multiple dosing at steady state, PKNCA won't calculate AUClast and other related parameters (understandable), but somehow aucint.last (with the interval pre-specified) is still estimated. I think the same rule should be applied to both auclast and aucint.last, so in this case, neither parameter cannot be calculated.

[billdenney]

I agree that the same rule should be applied. Can you please provide a reproducible example so that I can look into this unexpected result when no predose value is present?

[billdenney]

AUCint.last is calculated starting with rules related to AUClast for calculation. It will (almost) always return a value, the same as AUClast. It interpolates between time points that are above the LLOQ (as you likely expect) and its extrapolation after Tlast is all zero (see here for the description: https://humanpred.github.io/pknca/articles/v03-selection-of-calculation-intervals.html#parameters-available-for-calculation-in-an-interval).

My guess is that the parameter you're looking for would be either aucint.inf.pred or aucint.inf.obs.