## Setup in Phoenix/WinNonlin [Study Assessment]

Hi Angus,

» It is a very well known drug and many MR formulations are on the market. (You have experience with this drug).

If we are talking about the same goody: Watch out for polymorphism… Sometimes you have poor metabolizers in the study where enzymes get saturated at higher doses. In one subject I once got a slope of 1.54 for AUC and 2.16 for C

» There are only two does levels to plot. The relationship is as follows:

»

» LnPK=B0+B1*Ln(dose) where LnPK pertains to Cmax or AUC.

»

» So we have a regression line going through the points: we evaluate the slope (B1), intercept

OK, so far.

» and the confidence intervals about the slope to evaluate dose proportionality.

This is beyond me.

» Brian Smith in Pharm Research year 2000 has extended the approach from the original UK working party.

Yep. Smith

» It seems that you can calculate intrasubject and intersubject variance e.g. for AUC and partial AUC from this approach.

Correct.

» I do not follow how to do it. I use the usual intrasubject and intersubject values from Phoenix WinNonlin 6.4 and I am happy with that.

If you are happy with that, what is your question?

If you want to reproduce Smith’s results in Phoenix/WinNonlin: Start with a worksheet (columns subject, dose, Cmax, AUC, whatsoever). log-transform: dose, Cmax, … and weight=1/logCmax, … Send to LME. Map Subject as Classification, logCmax as Rgeressor, and logCmax as Dependent.

Model Specification: logCmax

Fixed Effects Confidence Level: 90%

Variance Structure / Random 1: Subject

With Smith’s C

0.7617 (90% CI: 0.6696, 0.8539), slightly different from the reported 0.7615 (0.679, 0.844). Why? Duno.

See also chapter 18.3 in Chow/Liu. Without explanation they recommend a 95% CI but a 90% CI in elaborating Smith’s approach. In general I prefer a weighted model (hence the transformation above). Fits much better.

PS: Can you ask “the other worker” why he/she calculated the 98% CI?

» It is a very well known drug and many MR formulations are on the market. (You have experience with this drug).

If we are talking about the same goody: Watch out for polymorphism… Sometimes you have poor metabolizers in the study where enzymes get saturated at higher doses. In one subject I once got a slope of 1.54 for AUC and 2.16 for C

_{max}over a only twofold dose range! For the other subjects it got 1.05 and 1.02 with a very narrow CI.» There are only two does levels to plot. The relationship is as follows:

»

» LnPK=B0+B1*Ln(dose) where LnPK pertains to Cmax or AUC.

»

» So we have a regression line going through the points: we evaluate the slope (B1), intercept

OK, so far.

» and the confidence intervals about the slope to evaluate dose proportionality.

This is beyond me.

*df*=*n*–*p*where*n*is the number of data points and*p*the number of parameters. How can you calculate a CI with*df*= 0?» Brian Smith in Pharm Research year 2000 has extended the approach from the original UK working party.

Yep. Smith

*et al.*use a mixed-effects model, where subjects are a random effect. Thus we increase*n*. Now a CI is possible even for*p*= 2.» It seems that you can calculate intrasubject and intersubject variance e.g. for AUC and partial AUC from this approach.

Correct.

» I do not follow how to do it. I use the usual intrasubject and intersubject values from Phoenix WinNonlin 6.4 and I am happy with that.

If you are happy with that, what is your question?

If you want to reproduce Smith’s results in Phoenix/WinNonlin: Start with a worksheet (columns subject, dose, Cmax, AUC, whatsoever). log-transform: dose, Cmax, … and weight=1/logCmax, … Send to LME. Map Subject as Classification, logCmax as Rgeressor, and logCmax as Dependent.

Model Specification: logCmax

Fixed Effects Confidence Level: 90%

Variance Structure / Random 1: Subject

With Smith’s C

_{max}-data of Table 1 I got for the slope:0.7617 (90% CI: 0.6696, 0.8539), slightly different from the reported 0.7615 (0.679, 0.844). Why? Duno.

(jitter added to doses)

See also chapter 18.3 in Chow/Liu. Without explanation they recommend a 95% CI but a 90% CI in elaborating Smith’s approach. In general I prefer a weighted model (hence the transformation above). Fits much better.

` SSQ AIC Var(Subject) Var(Res)`

w = 1 0.07389 9.247 0.1120 0.01456

w = 1/logCmax 0.01548 -8.917 0.1108 0.003076

PS: Can you ask “the other worker” why he/she calculated the 98% CI?

—

Helmut Schütz

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

Science Quotes

*Dif-tor heh smusma*🖖Helmut Schütz

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

Science Quotes

### Complete thread:

- Dose Proportionality and Variance AngusMcLean 2016-05-11 16:55 [Study Assessment]
- More information, please Helmut 2016-05-12 14:34
- More information, please AngusMcLean 2016-05-13 16:40
- Setup in Phoenix/WinNonlinHelmut 2016-05-14 02:26
- Setup in Phoenix/WinNonlin AngusMcLean 2016-05-14 18:54
- Setup in Phoenix/WinNonlin Helmut 2016-05-15 14:47
- Setup in Phoenix/WinNonlin AngusMcLean 2016-05-15 15:17
- Phoenix 64 Warning Helmut 2016-05-15 15:56
- Phoenix 64 Warning AngusMcLean 2016-05-15 20:11
- OT: imperial vs. metric units Helmut 2016-05-16 16:26

- Phoenix 64 Warning AngusMcLean 2016-05-15 20:11
- Setup in Phoenix/WinNonlin ElMaestro 2016-05-15 20:54
- Setup in Phoenix/WinNonlin AngusMcLean 2016-05-15 22:30

- Phoenix 64 Warning Helmut 2016-05-15 15:56

- Setup in Phoenix/WinNonlin AngusMcLean 2016-05-15 15:17

- Setup in Phoenix/WinNonlin Helmut 2016-05-15 14:47
- Setup in Phoenix/WinNonlin AngusMcLean 2016-05-16 21:00
- NCSS vs. PHX/WNL vs. SAS Helmut 2016-05-17 01:50
- NCSS vs. PHX/WNL vs. SAS - Validation? mittyri 2016-05-18 08:23
- Diagnostics ElMaestro 2016-05-18 09:20
- Diagnostics: R and Phoenix Helmut 2016-05-18 15:14
- Diagnostics: R zizou 2016-05-22 19:07
- Diagnostics: R Helmut 2016-05-23 01:22
- SASian potpourri d_labes 2016-05-24 12:02
- Compilation Helmut 2016-05-24 14:27
- REML or not d_labes 2016-05-24 16:33
- complete or not Helmut 2016-05-24 16:57

- Compilation AngusMcLean 2016-05-26 16:46
- doubts about NCSS Helmut 2016-05-26 19:13
- Doubts about NCSS zizou 2016-05-26 23:38

- doubts about NCSS Helmut 2016-05-26 19:13
- Compilation AngusMcLean 2016-05-28 00:51
- Kenward-Roger? Helmut 2016-05-28 15:59

- 90% confidence interval for R_dnm Shuanghe 2019-01-04 17:45
- 90% confidence interval for R_dnm d_labes 2019-01-05 14:01
- Visualizing lmer and limits mittyri 2019-01-06 17:00
- Visualizing lmer and limits Shuanghe 2019-01-07 11:05
- Visualizing lmer and limits d_labes 2019-01-07 15:08
- Visualizing lmer and limits mittyri 2019-01-13 23:53

- 90% confidence interval for R_dnm Shuanghe 2019-01-07 10:53
- 90% confidence interval for R_dnm d_labes 2019-01-07 15:17
- 90% confidence interval for R_dnm Shuanghe 2019-01-07 17:11
- 90% confidence interval for R_dnm d_labes 2019-01-07 18:24
- offtop: greek letters and tables mittyri 2019-01-08 00:19
- OT: greek letters and symbols Helmut 2019-02-02 16:04

- 90% confidence interval for R_dnm Shuanghe 2019-01-07 17:11

- 90% confidence interval for R_dnm d_labes 2019-01-07 15:17

- Visualizing lmer and limits mittyri 2019-01-06 17:00

- 90% confidence interval for R_dnm d_labes 2019-01-05 14:01

- REML or not d_labes 2016-05-24 16:33

- Compilation Helmut 2016-05-24 14:27

- SASian potpourri d_labes 2016-05-24 12:02

- Diagnostics: R Helmut 2016-05-23 01:22

- Diagnostics: R zizou 2016-05-22 19:07

- Diagnostics: R and Phoenix Helmut 2016-05-18 15:14
- Smith’s paper Helmut 2016-05-18 14:44
- Smith’s paper d_labes 2019-01-05 15:00

- Diagnostics ElMaestro 2016-05-18 09:20

- NCSS vs. PHX/WNL vs. SAS - Validation? mittyri 2016-05-18 08:23

- NCSS vs. PHX/WNL vs. SAS Helmut 2016-05-17 01:50

- Setup in Phoenix/WinNonlin AngusMcLean 2016-05-14 18:54

- Setup in Phoenix/WinNonlinHelmut 2016-05-14 02:26

- More information, please AngusMcLean 2016-05-13 16:40

- More information, please Helmut 2016-05-12 14:34