## Difference between actual and published PK parameters [Study As­sess­ment]

Hi Loky do,

» […] I have a question regarding dapoxetine biphasic half-life; the initial and terminal half-life (the initial half-life which is 1.3-1.5 hours and the terminal half-life which is 15-19 hours) does this biphasic nature affect the practical obtained half life of the drug?

I don’t understand what you mean by ‘practical obtained half life’. Do you mean the apparent terminal half life estimated from a lin/log-regression? Let’s explore the fasted study of Yan et al.:

Here we see a straight line starting at 12 hours. Therefore, we can reliably estimate $$\small{\lambda_\textrm{z}}$$.

Since you wrote in your original post
» » » the drug was detected […] for only 24 hours in most of the volunteers.
it might well be that distribution was not complete ≤ 12 hours and hence, the estimated elimination contaminated. When I make a rough estimation from the concentrations in the figure between 8 and 24 hours, I get a half life of ~9.6 hours. Not exactly yours, but close.

Possibly you see such patterns (+ – +) in the fits:

Of course, what ElMaestro wrote might be another explanation.

However, in BE we are interested in detecting potential differences in the absorption of formulations. Once absorption is complete,* anything else is a property of the drug and should not be a regulatory concern.

» and is this nature could affect the drug variability?

In BE we make the – rather strong – assumption that clearance is constant (background). If this is not the case (likely…), it will negatively affect the residual variability.
In a two-compartment system, we have three clearances: The total body clearance (associated with elimination) and two inter-compartment clearances (associated with distribution). If you are in church of volumes of distribution / rate-constants:$$\begin{matrix} \overset{k_\textrm{a}}{\longrightarrow} & \boxed{V_1} & \overset{k_{12}}{\underset{k_{21}}{\rightleftharpoons}} & \;\;\boxed{V_2}\\ & \phantom{0}\downarrow \tiny{k_\textrm{e}} & & \end{matrix}$$ In simple terms: We can expect that in a multicompartment system the between-occasion variability to be larger than in a one-compartment system.

• Speaking of IR formulations. Once we are crossing the Rubicon of flip-flop PK ($$\small{k_\textrm{a}\leq k_\textrm{e}}$$ is common for prolonged release formulations), we have to follow the profile much longer.

Dif-tor heh smusma 🖖
Helmut Schütz

The quality of responses received is directly proportional to the quality of the question asked. 🚮
Science Quotes