+--------------------------------------------------------------------+
 |                                                                    |
 |                 ENTERHEPATIC CIRCULATION EXAMPLES                  |
 |                                                                    |
 +--------------------------------------------------------------------+

 The  help item "Model Time examples" gives a fragment of code for mod-
 elling EHC (Enterohepatic Circulation, which is also called Enterohep-
 atic  Recycling).  It illustrates the use of MTIME parameters to model
 instantaneous changes in  differential  equations.   This  help  item,
 "Enterhepatic  circulation  examples",  describes two fully-worked out
 control streams in the examples directory.  They  demonstrate  how  to
 generalize the fragment to multiple sequential doses.

 Both  examples  use  the  same  data.  (hillss.dat and mtimess.dat are
 identical).  There is a steady-state bolus dose at time 0.  The inter-
 dose  interval  II  is  12,  and  there  are  enough  additional doses
 (ADDL=100) to continue the dosing pattern  throughout  the  data  set.
 There  are  "other"  records every 4 units till time 140 to allow com-
 partment amounts to be displayed and  there  is  a  final  observation
 record at time 144.

 mtimess.ctl
      This  example  incorporates  the fragment into a complete control
      stream.  MTIME parameters are used to turn on  and  off  the  EHC
      terms  in  the  differential  equations.   The variable FLAG is 1
      between times MTIME(1)  and MTIME(2) after each  dose  event  and
      turns  on  the  EHC terms.  After time MTIME(2) is reached, a new
      set of MTIME's is defined which affect the next dosing  interval.
      MTIME  parameters  are  not  dose-related  parameters and have no
      effect  on  steady-state  dose  events.   Even  if  PK   computes
      MTIME(i)<  II,  this  produces  future changes in the system, and
      does not apply retroactively to the preceding implied doses.
      (See Guide VI, Section V.F.4, Note 4).
      A steady-state dose record should not be used.  Instead,  the  SS
      dose  record is described as a transient dose with SS=DROP on the
      $INPUT record.

 hillss.ctl
      This control stream does not use MTIME. Instead,  a  smooth  step
      model using Hill terms in a sigmoid emax model is used.  The $DES
      code has to compute all the necessary variables.  Flag1 and flag2
      are  continuous  variables  that  change from 0 to 1 at the times
      corresponding to the MTIME's.  The FLAG variable  is  similar  to
      FLAG  in  mtimess.ctl.  The changes to the differential equations
      are not instantaneous, but they are continuous. If  the  exponent
      in  the  Hill term is made larger, the predictions approach those
      of the MTIME model.  However, very large values of  the  exponent
      can  lead  to numerical difficulties in PREDPP. Smaller values of
      the exponent may be more realistic  physiologically.   A  Steady-
      State dose event record is used with this model.

      Note  that  SS dose records should only be used when the kinetics
      implemented in the model coincides with the II (interdose  inter-
      val) of the SS record.  Just as MTIME's cannot affect the differ-
      ential equations retroactively, changes to the differential equa-
      tions  that  happen  in the future cannot affect the Steady-state
      calculations.  For example, with II=12, then the kinetics  should
      not  be  different in the interval of time 0 to 12 vs. time 12 to
      24 or time 24 to 36, etc.  In each  interval,  changes  occur  at
      theta(8)  and  theta(8)+theta(9) after the start of the interval.
      $DES computes the first change time for each interval  using  the
      INT function:
      mt1=II*INT(T/II)+theta(8)

      NM-TRAN  gives  a  warning  about  the use of the INT function in
      $DES:

  (WARNING  68) THE INT, MOD, MIN, OR MAX FUNCTION IS BEING  USED  OUT-
 SIDE OF A
  SIMULATION BLOCK. IF THE FUNCTION VALUE AFFECTS THE VALUE OF THE
  OBJECTIVE FUNCTION, THEN AN ERROR WILL PROBABLY OCCUR.

      This  warning  may  be  disregarded.  Discontinuties occur at the
      ends of the integration intervals, but  the  kinetics  are  unaf-
      fected.   For example, when T<12, the value of 12*INT(T/12) is 0.
      The values of FLAG1 and FLAG2 are initially 0 and FLAG is 0.   As
      T  approaches  the  end point T=12, the values of FLAG1 and FLAG2
      both become 1 and FLAG is 0.  At the end  point  when  T=12,  the
      value  of 12*INT(T/12) is 12 and both FLAG1 and FLAG2 are 0.  The
      discontinuity in mt1 and mt2 does not affect  the  FLAG  variable
      because  FLAG1 and FLAG2 are both 0 or both 1 in the neighborhood
      of the discontinuity, and the kinetics are continuous.

 In these examples, EHC is driven by the dose events.  The EHC  changes
 can also driven by the clock.  Suppose every 12 hours, a new EHC cycle
 begins.  (The value 12 is chosen so that the two  versions  will  give
 the same predictions.)  E.g., instead in mtimess.ctl, instead of
   MTIME(1) = MTIME(1)+II
  compute
   MTIME(1) = MTIME(1)+12

 In hillss.ctl, instead of
   mt1=inter*INT(T/inter)+theta(8)
  compute
   mt1=12*INT(T/12)+theta(8)

 There is no difference in the results.

REFERENCES: Guide VI Section III.F.9
REFERENCES: Guide VI Section V.F
REFERENCES: Guide IV Section V.C.5


  
Go to main index.
  
Created by nmhelp2html v. 1.0 written by Niclas Jonsson (Modified by AJB 5/2006,11/2007,10/2012)