Back to the Top
Just wondering what or if there is a consensus out in the academic world for
AUC dosing and Carboplatin when an elderly patient has a serum creatinine
less than 1 (say 0.5 to 0.7) this occurs not infrequently and if one uses
0.5 to 0.7 in the Cockcroft/Gault formula one gets a clearance approaching
100 mls/minute and an AUC dose that approaches the original 400mg/m2
carboplatin dosing. Doesn't this defeat the purpose of using AUC to lower
the dose and decrease toxicity.
Thanks
Patrick Skeffington, PharmD, MS
Elliot Hospital
One Elliot Way
Manchester, NH 03103
(603)663-7653
Back to the Top
Patrick,
An answer to your question might be considerd in two independent
parts - the prediction of creatinine clearance and the prediction of
carboplatin dose.
"Skeffington, Patrick (by way of David Bourne)" wrote:
> Just wondering what or if there is a consensus out in the academic world for
> AUC dosing and Carboplatin when an elderly patient has a serum creatinine
> less than 1 (say 0.5 to 0.7) this occurs not infrequently and if one uses
> 0.5 to 0.7 in the Cockcroft/Gault formula one gets a clearance approaching
> 100 mls/minute
Some insight into the properties of the C&G formula derives from
considering it as a model which uses age, weight and sex to predict
creatinine production rate (CPR). Then given a serum creatinine (Scr)
the creatinine clearance (CLcr) can be predicted by assuming steady
state:
Rate in = CPR
Rate out = Scr * CLcr
at steady state:
CPR=Scr * CLcr
CLcr = CPR/SCr
Age, weight and sex are covariates that predict muscle mass and thus
CPR. Because these covariates were derived from studying relatively
healthy subjects the influence of other covariates such as body
composition also need to be considered. In a patient with cancer the
ratio of muscle mass to total body weight may be lower because of the
wasting associated with a terminal illness. In that case one might
guess (by looking at the patient) that the CPR prediction should be
reduced by say 30% because of muscle loss. This would tnen lead to a
30% reduction in predicted CLCr. Conversely, a body builder with an
increased muscle mass to body weight ratio might need a 30% increase
in CPR.
Serum creatinine less than say e.g. 0.06 mmol/L (0.68 mg/dL), has
been shown to overpredict observed amninoglycoside clearance if a
linear relation between CLcr and AG clearance is assumed. This had
led some to advocate rounding up all SCr values less than 0.06 to
0.06 and using the rounded up value to predict AG clearance. This is
an empirical algorithm with no underlying theory. Its only
justification is that it seems to work better than using the original
Scr. It may have similar merits for predicting carboplatin clearance.
> and an AUC dose that approaches the original 400mg/m2
> carboplatin dosing. Doesn't this defeat the purpose of using AUC to lower
> the dose and decrease toxicity.
Assuming that one has a reliable estimate of creatinine clearance
(e.g. by using a radiolabelled renally cleared substance such as
CrEDTA,Iothalamate, or 99TCDTPA) then one can apply a model for
prediction of carboplatin clearance and thus the carboplatin dose to
achieve the target AUC (e.g. Calvert et al. J Clin Oncol
1989;7:1748-1756, Chatelut et al. JNCI 1995;87:573-579). If the CLcr
predicts a dose of 400 mg/m2 then this is a necessary consequence of
the models that relate CLcr to carboplatin clearance and thus to
carboplatin AUC. If you do not have a reliable method of predicting
Clcr then of course the carboplatin dose prediction will not be
reliable.
I do not agree with your assertion that the purpose of using AUC is
to lower the dose. The philosophy of the target concentration
approach is to adjust the dose, down or up, to achieve an AUC that is
most likely to be effective with acceptable toxicity. If you feel
that the target AUC in a patient should be lower (perhaps because of
previous or concomitant treatment which may incrased toxicity) then
it makes sense to use the CLcr to predict carboplatin clearance as
before but then use the carboplatin clearance with a lower target AUC
to predict a lower dose.
--
Nick Holford, Divn Pharmacology & Clinical Pharmacology
University of Auckland, 85 Park Rd, Private Bag 92019, Auckland, New Zealand
email:n.holford.-a-.auckland.ac.nz
http://www.health.auckland.ac.nz/pharmacology/staff/nholford/
Back to the Top
[Two replies - db]
From: Roger Jelliffe
Date: Tue, 02 Jul 2002 15:25:06 -0700
To: david.aaa.boomer.org
Subject: Re: PharmPK Re: Updated ODA consensus document.
Dear all:
About AUC dosing and CCR. Two things. One is that Cockroft and
Gault is good when serum creatinine is stable. But if it is changing from
day to day, for example, in an unstable patient, you might want to use the
method we developed first back in the early 70's, before C+G. You can get
an updated version of it on our web site www.lapk.org
Another thing is that we wonder if there has not been an
overconcentration on the serum AUC at the expense of other things. Most
drugs have significant tissue distribution, and differences in the exchange
rate constants between serum and nonserum compartments may have significant
consequences on drug therapy.
For example, take digoxin. Don't just use a 1 compartment model as
so many do. Use the very nice model of Reuning and colleagues, from back in
1973 (!! - why have so few cardiologists picked up on it??) where they show
a much better correlation between inotropic effect and concentrations in
the peripheral nonserum compartment than with serum concentrations. We find
the same thing clinically. There is a much better correlation between
peripheral nonserum concentrations than with serum concentrations (see our
clinical USC*PACK software). Now when one uses a 1 compartment model of
digoxin in the presence of quinidine, there is general agreement that the
clearance of digoxin is very significantly reduced. to 1/3 or 1/4, in the
presence of quinidine. If you look only at the serum AUC this is what you
would think. However, if you use a 2 compartment model, you see that the
REASON the serum concentrations are so high and the clearance thus so
apparently reduced is because the peripheral uptake of drug is so reduced,
as the quinidine reduces the tissue uptake of digoxin. Adding up both the
total amounts of drug in the serum and the peripheral compartments, there
is actually very little difference in the total amount of digoxin on the
body as a whole - it is mostly simply redistributed. If all you look at is
the serum concentration and the serum AUC, you will never see this. But if
you look at the peripheral concentration and its AUC, you will get a
totally different view of what is really going on.
The point here is that many drugs used in transplant and cancer
therapy have just such significant uptake in nonserum compartments. If all
we do is look at the serum AUC, we will miss a great deal of what is really
going on. Think about using better models - not just 1 compartment ones,
and look at the concentrations in other compartments and their AUC's, and
correlate them with patient behavior. You may well see a much richer view
of things, just as with digoxin, where the reduced peripheral compartment
concentrations and AUC's correlate much better with the reduced inotropic
effect of digoxin in the presence of quinidine than they do with the very
elevated serum concentrations and their AUC's.
Very best regards,
Roger Jelliffe
Roger W. Jelliffe, M.D. Professor of Medicine,
Laboratory of Applied Pharmacokinetics,
USC Keck School of Medicine
2250 Alcazar St, Los Angeles CA 90033, USA
email= jelliffe.-at-.hsc.usc.edu
Our web site= http://www.lapk.org
---
From: "Stephen Duffull"
Date: Wed, 3 Jul 2002 08:33:10 +1000
To: david.-a-.boomer.org
Subject: RE: PharmPK Re: Updated ODA consensus document.
Patrick,
> Just wondering what or if there is a consensus out in the
> academic world for
> AUC dosing and Carboplatin when an elderly patient has a serum creatinine
> less than 1 (say 0.5 to 0.7) this occurs not infrequently and if one uses
> 0.5 to 0.7 in the Cockcroft/Gault formula one gets a clearance approaching
> 100 mls/minute and an AUC dose that approaches the original 400mg/m2
> carboplatin dosing.
There is surprisingly little work done on this. In a paper in '97 in CCP we
used 0.06 (mmol/L) as out cut off (note: the change in units) with
carboplatin. Empirically we have found that very low values of serum
creatinine do not always correspond to very high values of creatinine
clearance as predicted by the Cockcroft and Gault eqn. This is probably due
to reduced production of creatinine rather than increased clearance. We
have verified this cut-off with other drugs such as gentamicin (BJCP '97
ish). I think that Kirkpatrick et al. BJCP 1999;47:637-43 provides a good
summary of the cut-off method - with a description of it for gentamicin.
But back to carboplatin. Other than a few predictive models developed using
popn software that use serum creatinine (or creatinine clearance) - the
majority of work seem to use isotopic determination of GFR (eg. Cr-EDTA or
Tm-DTPA) in conjunction with the Calvert eqn. The Calvert eqn does not
perform well with estimates of GFR using the Cockcroft and Gault method (see
a publication in CPK 97;33:161-183 that reviews this)...
I hope this helps
Steve
Stephen Duffull
School of Pharmacy
University of Queensland
Brisbane 4072
Australia
http://www.uq.edu.au/pharmacy/duffull.htm
Back to the Top
Patrick:
We recently reviewed our PK data on carboplatin. We focused on the
issue of obesity, but found that empiric modifications of actual vs
ideal did not make a large difference in the accuracy of the
resulting AUC vs intended. I would agree with Nick that the same is
likely with issues of low creatinine concentrations. A few
suggestions:
1. Consider whether the carboplatin is given with curative or
palliative intent. If curative, I would use the lower creatinine
yielding the higher carboplatin dose. If palliative, I would still
dose using the low creatinine value, but others may be more
comfortable decreasing the dose using the lower limit for creatinine
of 1 mg/dl (88.3 microM). Remember that, if the patient has minimal
toxicity, the risk is that they are underexposed to the drug and
subsequent escalations in dose may be warranted.
2. Try the Chatelut equation (or the Benezet-Chatelut equation for
obese patients using mean of actual and ideal weights) as a
comparator to the results from the CG equation.
3. Check out the new paper by JW Wright, et al, that provides what I
would argue is a new standard for estimating creatinine clearance
from body size, age, and serum creatinine, replacing Cockcroft Gault.
Instead of generating a regression curve from urine collections, they
used 51Cr-EDTA as the GFR benchmark, and even found that elevated
creatine kinase and assay methods impacted the regression. Equations
are given for each case. In our post hoc analysis, the
Benezet-Chatelut equation using mean weight gave best accuracy,
followed closely by the combination of Wright for GFR and Calvert for
dose. All others (including tweaking body weights for CG equation)
had prediction errors significantly different than 0.
References below:
Chatelut E, Canal P, Brunner V, et al. Prediction of carboplatin
clearance from standard morphological and biological patient
characteristics. J Natl Cancer Inst 87: 573-580, 1995.
Benezet S, Guimbaud R, Chatelut E, et al. How to predict carboplatin
clearance from standard morphological and biological characteristics
in obese patients. Ann Oncol 8: 607-609, 1997.
Wright JG, Boddy AV, Highley M, Fenwick J, McGill A, Calvert AH.
Estimation of glomerular filtration rate in cancer patients. Br J
Cancer 84: 452-459, 2001.
Take care,
Paul (from the "Six Flags over Wisconsin" School of Pharmacokinetics)
Paul Hutson, Pharm.D.
Associate Professor (CHS)
UW School of Pharmacy
NOTE NEW ADDRESS effective 6/2001
777 Highland Avenue
Madison, WI 53705-2222
Back to the Top
Patrick (and all)
Are you aware of any clinical papers in which the rounding-up of
serum creatinine to 1 mg/dl (or say, 90 uM) is proposed or validated?
I can't find any.
Thanks in advance.
Paul
Paul Hutson, Pharm.D.
Associate Professor (CHS)
UW School of Pharmacy
NOTE NEW ADDRESS effective 6/2001
777 Highland Avenue
Madison, WI 53705-2222
Back to the Top
Dear Paul,
For its proposal, see an article by McCormack et al. Simple approach to
dosage adjustment in patients with renal impairment. Am J Health Syst Pharm
1997;54:2505-2509.
Additional information can be found in the upcoming letter to the editor, by
Oo et al.,Change in creatinine clearance with advancing age, Journal of
American Geriatric Society 2002, 50 (9):1-2
Kind regards,
Charles
Charles Oo, PharmD, PhD
Hoffmann-La Roche Inc.
Clinical Pharmacology
340, Kingsland Street
Nutley, NJ 07110-1199
Back to the Top
Paul
> Are you aware of any clinical papers in which the rounding-up of
> serum creatinine to 1 mg/dl (or say, 90 uM) is proposed or validated?
I have seen round up of SCr to 80 uM and to 60 uM (and various values
between these 2), but not 90uM.
Steve
Stephen Duffull
School of Pharmacy
University of Queensland
Brisbane 4072
Australia
http://www.uq.edu.au/pharmacy/duffull.htm
Back to the Top
Paul,
We did an "empirical" comparison of rounding up to 60, 70 or 88.4
umol/L (1g/dl) and found 60 or 70 were better than 88. Ref - Rosario
et al, Br J Clin Pharmacol 1988 46: 229-236.
The 1 mg/dl was suggested by Gral et al, J Am Geriatr Soc 1980 28:
492-6 and Zarowitz et al Ann Pharmacother 1992 26: 1205, 1210.
Alison
PharmPK Discussion List Archive Index page
Copyright 1995-2010 David W. A. Bourne (david@boomer.org)