Cumulation detection [PK / PD]
Hi BE-proff,
Not sure what you mean. According to the “Superposition Principle” linear pharmacokinetics could be tested by a comparison of AUC0-τ (steady state) with AUC0-∞ (single dose). Generally this is done in a study where after the single dose profile the drug is administered until steady state is reached.*
![[image]](img/uploaded/image355.png)
The comparison is done by a paired test – which assumes no period effects. I have never seen a crossover in 35 years… Would be a logistic nightmare.
If the 90% CI of AUC0-τ/AUC0-∞ is outside the acceptance range, nonlinear PK is proven. See also this lengthy thread.
The superposition principle is applicable to AUCs only (that’s why Friedrich Hartmut Dost called it “Gesetz der korrespondierenden Flächen” – Law of Corresponding Areas back in 1953).
Comparisons of Css,max/Cmax and tss,max–tmax are possible but of doubtful value.
Whichever you might think of.
A common one is the “Accumulation Index”:
\(R=\frac{1}{1-2^\epsilon}\), where \(\epsilon=\tau/t_{^1/_2}\).
Phoenix/WinNonlin uses a different formula
\(R=\frac{1}{1-\exp(-\lambda_\text{z}\cdot \tau)}\), which gives the same result (homework: why?).
\(\small{R}\) gives you an idea how much higher average concentrations in steady state are if compared to a single dose. Example: Half-life 12 h (\(\small{\lambda_\text{z}}\) 0.05776) and dosing interval 24 h \(\small{\Rightarrow R\;1.\dot{3}}\) (i.e., concentrations in steady state will be ⅓ higher than after a single dose). If you decrease the dosing interval to \(\small{t_{\small{^1/_2}}}\), \(\small{R}\) will be 2.
I’m not a big fan of R since it might be difficult to obtain a reasonably good estimate of λz / t½ in steady state. In the BE-context we don’t sample beyond τ. If you are interested in a good estimate of λz (e.g., in phase I studies of new drugs) you might sample longer.
Furthermore, R is strictly valid only for the one-compartment model. For more compartments it is approximate (since based on the elimination phase only).
❝ Is it possible to detect cumulation of a substance using "classic" PK-parameters (AUC, Cmax, Tmax, etc)?
Not sure what you mean. According to the “Superposition Principle” linear pharmacokinetics could be tested by a comparison of AUC0-τ (steady state) with AUC0-∞ (single dose). Generally this is done in a study where after the single dose profile the drug is administered until steady state is reached.*
![[image]](img/uploaded/image355.png)
The comparison is done by a paired test – which assumes no period effects. I have never seen a crossover in 35 years… Would be a logistic nightmare.
If the 90% CI of AUC0-τ/AUC0-∞ is outside the acceptance range, nonlinear PK is proven. See also this lengthy thread.
The superposition principle is applicable to AUCs only (that’s why Friedrich Hartmut Dost called it “Gesetz der korrespondierenden Flächen” – Law of Corresponding Areas back in 1953).
Comparisons of Css,max/Cmax and tss,max–tmax are possible but of doubtful value.
❝ Maybe additional metrics to be calculated?
Whichever you might think of.

\(R=\frac{1}{1-2^\epsilon}\), where \(\epsilon=\tau/t_{^1/_2}\).
Phoenix/WinNonlin uses a different formula
\(R=\frac{1}{1-\exp(-\lambda_\text{z}\cdot \tau)}\), which gives the same result (homework: why?).
\(\small{R}\) gives you an idea how much higher average concentrations in steady state are if compared to a single dose. Example: Half-life 12 h (\(\small{\lambda_\text{z}}\) 0.05776) and dosing interval 24 h \(\small{\Rightarrow R\;1.\dot{3}}\) (i.e., concentrations in steady state will be ⅓ higher than after a single dose). If you decrease the dosing interval to \(\small{t_{\small{^1/_2}}}\), \(\small{R}\) will be 2.
I’m not a big fan of R since it might be difficult to obtain a reasonably good estimate of λz / t½ in steady state. In the BE-context we don’t sample beyond τ. If you are interested in a good estimate of λz (e.g., in phase I studies of new drugs) you might sample longer.
Furthermore, R is strictly valid only for the one-compartment model. For more compartments it is approximate (since based on the elimination phase only).
- “True” steady state is reached at t = ∞. Too long for us mortals. However, most people are happy by administering for five to seven half-lives. IIRC, only ANVISA requires ten half-lives.
n % ss
1 50.00
2 75.00
3 87.50
4 93.75
5 96.88
6 98.44
7 99.22
8 99.61
9 99.80
10 99.90
11 99.95
12 99.98
13 99.99
Formula: % ss = 100(1 – ½n )
—
Dif-tor heh smusma 🖖🏼 Довге життя Україна!![[image]](https://static.bebac.at/pics/Blue_and_yellow_ribbon_UA.png)
Helmut Schütz
![[image]](https://static.bebac.at/img/CC by.png)
The quality of responses received is directly proportional to the quality of the question asked. 🚮
Science Quotes
Dif-tor heh smusma 🖖🏼 Довге життя Україна!
![[image]](https://static.bebac.at/pics/Blue_and_yellow_ribbon_UA.png)
Helmut Schütz
![[image]](https://static.bebac.at/img/CC by.png)
The quality of responses received is directly proportional to the quality of the question asked. 🚮
Science Quotes
Complete thread:
- Cumulation detection BE-proff 2015-11-24 08:13 [PK / PD]
- Cumulation detectionHelmut 2015-11-24 13:41