NONMEM Users Network Archive

Hosted by Cognigen

RE: [NMusers] unbalanced data set

From: Michael Fossler <>
Date: Wed, 6 Jan 2016 14:55:35 +0000

At the risk of being tiresome about this topic, absent specific differences=
 between Phase 1 and Phase 2/3 data , e.g., renal function due to age or di=
sease states, etc., I'd argue that most of the differences seen between Pha=
se 1 and Phase 2/3 data are due to adherence. In a sense, then, much of the=
 differences in PK between these two groups is artificial, and due to the f=
act that patients do not reliably take their medication as prescribed, as o=
pposed to Phase 1 volunteers, where adherence is near 100%. Bernard Vrijens=
 has published a lot on this topic as it relates to PPK analyses. We, as a =
discipline, need to start pushing hard for adherence measures in clinical t=

As an n=1 case study , a few years ago, I was involved with an analysis o=
f a large Phase 2 study which consisted of an in-house phase, followed by d=
ischarge to home and an out-patient phase. The patients were significantly =
older and sicker than Phase 1 volunteers, so one might expect some PK diffe=
rences. When we analyzed the data from the in-house portion of the study, w=
e got results nearly identical to Phase 1. However, when we added in the ou=
t-patient phase, IIV on many of the parameters increased dramatically, and =
the residual error became extremely large. Clearly, patients were not takin=
g their medication as prescribed ( and as they wrote in their patient diari=
es). We ended up not using the out-patient portion of the data, which repre=
sents a huge waste of resources.

This irritates people when I say this, but we as a discipline are so enamor=
ed of finding that magical covariate(s) which will explain variability, but=
 we neglect the most important one of all: Did they take the medicine when =
they say they did? No biological covariate can have as big of an effect as =
adherence. Accounting for adherence routinely results in up to a 50% decrea=
se in residual variability - few standard covariates have this effect.

Fossler M.J. Commentary: Patient Adherence: Clinical Pharmacology's Embarra=
ssing Relative. Journal of Clinical Pharmacology (2015) 55(4): 365-367.

Michael J. Fossler, Pharm. D., Ph. D., F.C.P.
VP, Quantitative Sciences
Trevena, Inc<>
Office: 610-354-8840, ext. 249
Cell: 610-329-6636

From: [] On=
 Behalf Of Denney, William S.
Sent: Wednesday, January 06, 2016 8:33 AM
To: <>
Cc: Zheng Liu;
Subject: Re: [NMusers] unbalanced data set

Hi Zheng,

I'll take an intermediate view between Joachim and Nick.

The rich data from Phase 1 provides the ability to define the structural mo=
del and a few of the important covariates. The control of Phase 1 gives pr=
ecision that cannot be achieved in Phase 2 or 3 studies. But, there are us=
ually important differences between Phase 1 and later phase populations tha=
t makes the later phase separately important.

With later phase trials, the range of covariates is expanded [1]. On top o=
f the expanded covariate range, sometimes late-phase patient populations ar=
e categorically different than early phase [2].

In practice, this means that I fit a single model to all data. The model w=
ill allow for the dense data from Phase 1 with more inter-individual variab=
ility (IIV) terms (fix the IIV to 0 for sparse data) and the expanded covar=
iate range with a richer set of fixed effects as the model is expanded for =
later phase. Finally, due to typical differences in data quality, I will o=
ften include a different residual error structure for sparse data. This ap=
proach allows the complexity of the Phase 1 structural model to carry into =
the richness of the late phase covariate model.

[1] A specific example is that typically renal function is allowed to be lo=
wer especially when Phase 1 is in healthy subjects.
[2] My true belief is that there may be unobserved covariates causing what =
appears to be a categorical difference. The functional impact of that beli=
ef is semantic only. In practice, the model would include a categorical pa=



On Jan 6, 2016, at 4:09, "Joachim Grevel" <<mailto:jgr=>> wrote:
Dear Zheng,
This is indeed a fundamental and recurring problem in drug development. You=
 have rich data from Phase 1 studies (single ascending dose, multiple ascen=
ding dose, others e.g. QTc) and sparse data from Phase 3 studies. Should yo=
u mix them all in one large meta-analysis and derive the definitive popPK m=
odel for that drug/project?
After years of experience, I tend to not mix Phase 1 with Phase 3 data. Pha=
se 1 can be used to establish the first popPK model which may contain speci=
al features such as nonlinearities/saturation effects as a consequence of t=
he wide range of doses studied. This can be the starting point for the buil=
ding of a fit-for purpose model using Phase 3 data only. I have come to bel=
ieve that the specific patient population(s) of Phase 3 require their own p=
opPK model that predicts exposure without bias. This is then used in the ex=
posure-response (E-R) modelling that is important for market approval. Only=
 a dedicated Phase 3 popPK model, that does not carry unnecessary legacies =
of Phase 1 development, is fit for E-R modelling and can give the important=
 answers about the dose rate(s) to be put in the drug label.

I would be interested to hear some other opinions.

Good luck,

Joachim Grevel, PhD
Scientific Director
BAST Inc Limited
Science & Enterprise Park
Loughborough University
Loughborough, LE11 3AQ
United Kingdom

Tel: +44 (0)1509 222908<

From:<> [ma=] On Behalf Of Zheng Liu
Sent: 06 January 2016 02:03
Subject: [NMusers] unbalanced data set

Dear all,

I recently have a data set for pk parameters fitting. The issue is some pat=
ients have far more measurement points than others (i.e. a few patients hav=
e ~15 points, other patients have only 1 or 2). I speculate in the fitted p=
arameters, those patients with many points would contribute much more than =
those with less points. Then the population "average" values of fitted pk p=
arameters are not anymore average from all the patients, but more biased to=
 those patients with many points. This is not what I expect.

Of course I could take away some points from the patients with many points,=
 in order to be comparable to less-points patients. Then I will be forced =
to lose some information from the data set. I just wonder are there anyone =
who have better proposal to solve this problem? I appreciate your help very=

Best regards,



Notice: This e-mail message, together with any attachments, contains inform=
ation of Trevena, Inc., 1018 West 8th Avenue, King of Prussia, PA 19406, US=
A. This information may be confidential, proprietary, copyrighted and/or le=
gally privileged.
It is intended solely for use by the individual or entity named on this mes=
sage. If you are not the intended recipient, and have received this message=
 in error, please notify us immediately and delete it and any attachments f=
rom your system.

Received on Wed Jan 06 2016 - 09:55:35 EST

The NONMEM Users Network is maintained by ICON plc. Requests to subscribe to the network should be sent to: Once subscribed, you may contribute to the discussion by emailing: