Other thread subtopics:

Centering Covariates
Covariate Models Using Weight (Allometric Scaling)
Covariate Models Using CrCL

**********************************************************************

From: "Stephen Duffull" <sduffull@fs1.pa.man.ac.uk>
Subject: Re: Covariate Models Using Weight
Date: Mon, 22 Nov 1999 17:04:44 -0000

James Gallo wrote (in relation to a comment from Nick):

> I feel you also are missing my point about covariate modeling related to
> renal function or any other mechanism related to drug disposition. For
> some reason you seem to want to adhere to PRE-DEFINED Formulas to describe
> CRCL and subsequently drug clearance.

This was my original point. If there is little difference in the fit between two models (one "predefined" and one not) to the same data then incorporating the "predefined" model which has proven generality seems more mechanistic than developing a "context-sensitive" model empirically (which has unknown generality). The choice of the "predefined" model (or "context-insensitive" model) is then up to the modeller (obviously no one model is going to solve all problems).

Steve
=====================
Stephen Duffull
School of Pharmacy
University of Manchester
Manchester, M13 9PL, UK
Ph +44 161 275 2355
Fax +44 161 275 2396

 

 

*****

 

 

Date: Mon, 22 Nov 1999 13:42:12 -0600
From: James Gallo <JM_Gallo@fccc.edu>
Subject: Re: Covariate Models Using Weight

Steve,

I disagree. I don't see how one can consider "PRE-Defined" models as mechanistic and "context-sensitive" [rather poor terminology] as empirical. Either type of model makes use of the same type of data [concentration-time or urine data and a mix of covairates], and it is only in how the final model shakes-out that one could attribute it to be 'mechanistic' or not. My concern is that the PRE-DEFINED models may be a priori applied as a surrogate for drug clearance when such application is unnecessary given a unique dataset for the drug of interest. I'd also be cautious of characterizing PRE-DEFINED models as being of "proven generality". Also, who cares if the so-called "context-sensitive" model is of "unknown generality" as long as it can most accurately predict clearance [for instance] for Drug X.

Your last point that ultimately how one proceeds is up to the modeler is a good one.

jim gallo

 

 

*****

 

 

Date: Mon, 22 Nov 1999 12:02:11 -0800
From: LSheiner <lewis@c255.ucsf.edu>
Subject: Re: Covariate Models Using Weight

Seems to me that this whole controversy is settled be acting like a Bayesian: the "pre-defined" scientifically based model has a certain prior credibility, which can be stated using an informative prior distribution on, for example, the allometric exponent. Then, if the data really favor some other value, the exponent will adjust; if they do not, it will remain near it's prior.

Almost always, controversies like this can be resolved by choosing a "model" (in this case the full Bayesian framework) which is a superset of both of the special cases that find themselves in conflict ...

LBS.
--
Lewis B Sheiner, MD Professor: Lab. Med., Biopharm. Sci., Med.
Box 0626 voice: 415 476 1965
UCSF, SF, CA fax: 415 476 2796
94143-0626 email: lewis@c255.ucsf.edu

 

 

*****

 

 

Date: Tue, 23 Nov 1999 09:34:45 +1300
From: Nick Holford <n.holford@auckland.ac.nz>
Subject: Re: Covariate Models Using Weight

James Gallo wrote:

> Nick,
>
> I feel you also are missing my point about covariate modeling related to
> renal function or any other mechanism related to drug disposition. For
> some reason you seem to want to adhere to PRE-DEFINED Formulas to describe
> CRCL and subsequently drug clearance.

PRE-DEFINED = prior knowledge

As Lewis Sheiner has pointed out we can always test using a Bayesian formulation if our data really has something to add to what we know already.

I like to use prior knowledge if possible because in information about covariate relationships is usually sparse.

> Given the problem of developing a population model for DRUG X that undergoes
> appreciable renal clearance, I believe it is more rational to start from
> scratch.

That is your choice. I prefer to lean on all the scientific knowledge we have accumulated to date rather than reinvent the wheel.

> the task was to model a drug that underwent 100% hepatic clearance. One
> would not have [or want] PRE-DEFINED formulas to relate covariates and
> clearance, and would start from scratch to identify significant covariate-CL
> relationships.

It seems you have not read the earlier portion of this thread in which I specifically pointed out the existence of a "pre-defined" allometric model which uses weight as the covariate to predict clearance such as hepatic clearance. Once again I prefer to rely on biology to help me understand rather than simply describe.
--
Nick Holford, Dept Pharmacology & Clinical Pharmacology
University of Auckland, Private Bag 92019, Auckland, New Zealand
email:n.holford@auckland.ac.nz tel:+64(9)373-7599x6730 fax:373-7556
http://www.phm.auckland.ac.nz/Staff/NHolford/nholford.htm

 

 

*****

 

 

Date: Mon, 22 Nov 1999 23:55:10 +0100
From: Pierre Maitre <pmaitre@freesurf.ch>
Subject: Re: Covariate Models Using Weight

>Nick Holford wrote:
>>
>> James Gallo wrote:
>>>
>>> James Wright wrote:
>>>
>>>> "Piotrovskij, Vladimir [JanBe]" wrote: ....

....May I too?

Dear respected scientists

May a modest clinician add a word of philosophy to your religious debate?

You have been discussing the way body weight influences the elimination clearance and I liked this hot discussion very much. It is indeed important to understand the relationship between weight and elimination. I am just asking myself weither it is appropriate to plug into a pharmacokinetic model a formula like
Males CLcr = (143.5 - (1.095*Age) * Wt * 0.07
Females CLCr = (119.5 - (0.915*Age) * Wt * 0.07
or any other equation you want, without some evidence in your data to support your choice.

My NONMEM mentor, Sam Vozeh, used to tell me, back in 1984, that nonlinear regression is an art, not a science. So please allow me tu use unscientific words like " I believe". I believe that clinicians need that we tell them the average pharmacokinetics of a drug, the size of the inter-individual variability, and the influence of important covariates (and among them, possibly, body weight). "Important" is the keyword here and is the hardest to define. The size of "Important" has first to be compared with the size of the interpatient variability for the parameter in question. Who cares about a small 8% change of clearance due to a particular covariate, if the remaining inter-individual variability is 40% for that parameter? ... forget this covariate, it won't help the clinician. For a clinician, "important" would mean changing the dosing by at least 20% (this is not science, again, but my experience as an anesthesiologist administering drugs every day). How do we sort out the "important" covariates from those who are no important? The P value doesn't help in this matter: as you all know, statistically significant does not mean important. My answer to this question might be: what you see in your data is worth modeling, what you don't see in your data is not worth modeling. Graphical methods based on the plots of etas vs covariates are a good start (Xpose can be used for this purpose) and allow one to pick up the "Important" covariates and to find a simple model that would fit the data. Dosing recommendations for the clinician must be kept simple in order to be safe. The above equations are certainly very exact, but they are just too complex to be used at bedside. And their complexity give to the clinician a false sense of scientific truth (that the clinician translates into "precise prediction") whereas the prediction of the concentration is very vague and unprecise, due to the cloud of interpatient variability. To conclude, my point would be: for your model to be useful, keep is simple.

Pierre Maitre
Genolier / Geneva

 

 

 

*****

 

 

From: "Piotrovskij, Vladimir [JanBe]" <VPIOTROV@janbe.jnj.com>

To: nmusers@c255.ucsf.edu

Subject: RE: Covariate Models Using Weight

Date: Tue, 23 Nov 1999 08:41:03 +0100

MIME-Version: 1.0

 

I agree that having parameters of pre-defined formulas open for iterations

may improve the fit, and this is perhaps a reasonable approach if a drug is

eliminated exclusively via passive glomerular filtration. However, if the renal clearance is only a part of the total clearance, and the enzyme-dependent part may be affected by the same covariates (AGE, WT, etc.) differently, this approach will not work or will give misleading results.

Vladimir
----------------------------------------------------------------------
Vladimir Piotrovsky, Ph.D.
Janssen Research Foundation
Clinical Pharmacokinetics
B-2340 Beerse
Belgium
Email: vpiotrov@janbe.jnj.com

 

 

*****

 

 

Date: Tue, 23 Nov 1999 09:14:00 +0000
From: James <J.G.Wright@ncl.ac.uk>
Subject: Re: Covariate Models Using Weight

Dear nmusers,

I think everyone in this (extremely long) thread believes in using prior knowledge in some way. The problem is that this is inevitably a subjective process and having done some work with renal function in cancer patients I can safely say that none of the existing formula work terribly well in this population. As such, I don't want to use them (because of population specificity, problems with the assay method etc.). If you happen to be working in a population you consider exchangeable with that used in the original study then maybe you could use their information, but the way to do this may well be with prior that acknowledges uncertainty. However I can here the frequentists calling if you have a big sample the prior doesn't matter and if you have a little sample you will get biased results - it depends on the quality of your prior information.

To return to Steve's original question - if both fit the model equally, then this probably isn't a big issue (not that you'd believe it). Personally I would stick to the covariates which come from a clearly defined source, rather than putting age & weight in a predefined formula and then interpreting this to mean their effects on renal function are completely accounted for.

James

PS Is this the longest thread ever on the nmuser list?

 

 

*****

 

 

From: "Piotrovskij, Vladimir [JanBe]" <VPIOTROV@janbe.jnj.com>
Subject: RE: Covariate Models Using Weight
Date: Tue, 23 Nov 1999 11:36:22 +0100

I agree, it is already too long, but still, one more reason for not to rely very much on SCR and CLCR (irrespective to the formula used). Creatinine is excreted partly via an active process, tubular secretion, and its steady-state level may be affected by some drugs. One example I have found is given at the end of my mail. I presume there are more recent examples.

Vladimir
----------------------------------------------------------------------
Vladimir Piotrovsky, Ph.D.
Janssen Research Foundation
Clinical Pharmacokinetics
B-2340 Beerse
Belgium
Email: vpiotrov@janbe.jnj.com

++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++

Authors
Cahen R. Martin A. Francois B. Baltassat P. Louisot P.

Institution
Division of Nephrology, Lyon-South University Medical Center, Pierre Benite,
France.

Title
Creatinine metabolism impairment by an anticonvulsant drug, phenacemide.

Source
Annals of Pharmacotherapy. 28(1):49-51, 1994 Jan.

MeSH Subject Headings:
Adult
*Anticonvulsants/ae [Adverse Effects]Carbamazepine/tu [Therapeutic Use]
Case Report
*Creatinine/bl [Blood]
Epilepsy, Temporal Lobe/bl [Blood]
Epilepsy, Temporal Lobe/dt [Drug Therapy]
Glomerular Filtration Rate
Human
Male
Phenobarbital/tu [Therapeutic Use]
*Urea/aa [Analogs & Derivatives]
Urea/ae [Adverse Effects]

Abstract:

OBJECTIVE: To report two cases of increased true serum creatinine (Scr) without renal failure caused by an anticonvulsant drug, phenacemide, and to discuss the possible mechanisms. CASE SUMMARY: Two patients treated with phenacemide were investigated for markedly increased Scr and decreased creatinine clearance (Clcr) values. Glomerular filtration rates, as determined by 125I-iothalamate clearance, were normal in both patients and analytical interferences with the Jaffe reaction were excluded. After discontinuation of the drug, phenacemide concentrations became undetectable within 2 days but it took 7-14 days for Scr and Clcr to return to normal values. DISCUSSION: The Scr increase with phenacemide (120-170 percent) was higher than that reported with cimetidine or trimethoprim (10-40 percent) and could not be explained solely by inhibition of the tubular secretion of creatinine. The hypothesis of an overproduction of creatinine caused by phenacemide was ruled out by experimental studies in rats. Creatinine increase in tissues was lower than that in the serum of rats given phenacemide. In vitro creatinine influx into red blood cells was inhibited in a dose-dependent way by phenacemide. CONCLUSIONS: Increased Scr concentrations in these patients could be related to an inhibition of transport and a decrease in creatinine volume of distribution. Creatinine concentrations should not be considered when dosage adjustments of renally eliminated drugs are being calculated for patients with such metabolic interferences.

 

**********************************************************************

 

Other thread subtopics:

Centering Covariates
Covariate Models Using Weight (Allometric Scaling)
Covariate Models Using CrCL