NONMEM Users Network Archive

Hosted by Cognigen

Re: Absorption delay NOT completed before the next dose is given

From: Alison Boeckmann <alisonboeckmann>
Date: Thu, 12 Apr 2007 16:16:55 UT

The Users Guide and NM help are incorrect.
When the doses are transient (as oppposed to steady-state) there is no
such restriction
in PREDPP.
I am attaching a brief memo giving more details.

On Sat, 31 Mar 2007 18:04:01 -0400, "Samtani, Mahesh [PRDUS]"
<MSamtani
> Dear NM users,
>
> NM help files says the following about ALAG
>
>
>
> "When additional doses are specified on a dose event record, the
> absorption
> lag time applies to the dose and to all the additional doses. In this
> case
> the lag time should not exceed the (length of the) interdose interval"
>
>
>
> My questions are
> a) Why doesn't ADDL work for the above specific case?
> b) I am running linear models with ADVAN2 and ADVAN5 and absorption
> delay
> is NOT completed before the next dose is given. Due to the problem stated
> above I am not using ADDL. Instead I have a dosing record at every time a
> dose enters the depot compartment. Will this circumvent the problem
> stated
> above? In other words does ALAG always have to be smaller than II and
> why?
>
>
>
> Please advice&#8230;MNS
>
>
>
>
>
>
>
>
>
--
  Alison Boeckmann
  alisonboeckmann



Mahesh Samtani raised two questions about the restrictions on ALAG and II
described in the NONMEM Users Guides as follows.
--------
From Guide VI (1992) Chapter III.F.6:
When additional doses are specified on a dose event record (see section
V.K), the absorption lag time acting at the time on the dose record
applies to the dose and to all the additional doses. In this case the
lag time should not exceed the (length of the) interdose interval.
With a steady-state multiple dose (see section V.F) the absorption lag
time applies not only to this dose, but also to all the preceding
implied doses. With such a dose, the lag time should not exceed the
interdose interval.
--------
From on-line help (approx. 1996) "ABSORPTION LAG PARAMETER":
When additional doses are specified on a dose event record, the
absorption lag time applies to the dose and to all the additional
doses. In this case the lag time should not exceed the (length of the)
interdose interval. With a steady-state multiple dose the absorption
lag time applies not only to this dose, but also to all the preceding
implied doses. With such a dose, the lag time should not exceed the
interdose interval.
--------
>a) Why doesn't ADDL work for the above specific case?
 
The first issue is why the restriction (ALAG<=II) exists for
additional transient doses. In fact, there is no such restriction in
PREDPP. It appears that a mistake was made in Guide VI and that this
mistake was perpetuated in the on-line help entry.

The similar restriction for steady state doses does exist and can be
justified. The message from PREDPP is:

PK PARAMETER FOR ABSORPTION LAG IS GREATER THAN OR EQUAL TO STEADY
STATE DOSE INTERVAL

(Notice that this is a little stronger than what is in the Users
Guides, in that it does not allow ALAG=II)

With steady state and multiple doses, PREDPP computes the state vector
at the time t on the dose event record. If absorption lag is present,
PREDPP then advances the system by the amount of time given by ALAG.
If this time is greater than II, then there should be another dose
given during the advance, at t+II. But this would contradict the
principle that the SS dose is the LAST of a series of implied doses.
In general, it is the users responsibility to use the ADDL data item
(or to provide transient dose records) to continue the steady state
pattern into the future. PREDPP will not supply additional doses, not
even for the purpose of computing steady state levels.

The Users Guides will be corrected at some point in the future.
For now, the on-line help entry should be:

When additional doses are specified on a dose event record, the
absorption lag time applies to the dose and to all the additional
doses. With a steady-state multiple dose the absorption lag time
applies not only to this dose, but also to all the preceding implied
doses. With a steady-state dose, the lag time should be less than
the interdose interval.

> b) I am running linear models with ADVAN2 and ADVAN5 and absorption
> delay is NOT completed before the next dose is given. Due to the
> problem stated above I am not using ADDL. Instead I have a dosing
> record at every time a dose enters the depot compartment. Will this
> circumvent the problem stated above? In other words does ALAG always
> have to be smaller than II and why?

 
There is no restriction in PREDPP that absorption delay for a given
dose event record must be completed before a new dose is entered in to
the system.
Received on Thu Apr 12 2007 - 12:16:55 EDT

The NONMEM Users Network is maintained by ICON plc. Requests to subscribe to the network should be sent to: nmusers-request@iconplc.com.

Once subscribed, you may contribute to the discussion by emailing: nmusers@globomaxnm.com.