Individualizing Dosage Regimens: Learning about our pa8ents - - PowerPoint PPT Presentation

individualizing dosage regimens learning about our
SMART_READER_LITE
LIVE PREVIEW

Individualizing Dosage Regimens: Learning about our pa8ents - - PowerPoint PPT Presentation

Individualizing Dosage Regimens: Learning about our pa8ents op8mally while trea8ng them at the same 8me, achieving maximum precision throughout. Roger W.


slide-1
SLIDE 1

Individualizing ¡Dosage ¡Regimens: ¡Learning ¡ about ¡our ¡pa8ents ¡op8mally ¡while ¡trea8ng ¡ them ¡at ¡the ¡same ¡8me, ¡achieving ¡maximum ¡ precision ¡throughout. ¡ ¡

Roger W. Jelliffe, M.D. Professor of Medicine Emeritus, Founder and Director Emeritus, Laboratory of Applied Pharmacokinetics, USC Keck School of Medicine Children’s Hospital of Los Angeles Los Angeles, CA, USA jelliffe@usc.edu www.lapk.org

slide-2
SLIDE 2

Our ¡Lab ¡

Fearless ¡Founder ¡ Alan ¡Schumitzky ¡ Mike ¡Van ¡Guilder ¡ Aida ¡Bustad ¡ Dave ¡Bayard ¡ Michael ¡Neely ¡– ¡Hot ¡new ¡blood! ¡

Supported in part by NIH grants EB005803, GM 068968, and HD070886

slide-3
SLIDE 3

Tasks ¡to ¡be ¡done ¡for ¡each ¡pa8ent ¡ ¡

  • 1 ¡– ¡First, ¡learn ¡op8mally ¡about ¡the ¡drug ¡
  • 2 ¡– ¡Then, ¡op8mize ¡the ¡ini8al ¡dosage ¡regimen ¡
  • 3 ¡– ¡Then, ¡learn ¡about ¡the ¡drug ¡in ¡that

¡pa8ent ¡op8mally ¡while ¡you ¡must ¡treat ¡ ¡him/her ¡at ¡the ¡same ¡8me ¡

  • 4 ¡– ¡Con8nue ¡learning ¡and ¡trea8ng ¡op8mally ¡

¡throughout ¡the ¡course ¡of ¡treatment. ¡

  • 5 ¡– ¡There ¡are ¡tools ¡for ¡each ¡part ¡of ¡this ¡task. ¡
slide-4
SLIDE 4

Tool #1 - Nonparametric (NP-UPD) PK/PD models ¡

  • The ¡most ¡likely ¡(ML) ¡popula8on ¡parameter ¡

distribu8ons, ¡given ¡the ¡data, ¡“CAN ¡BE ¡ FOUND” ¡in ¡a ¡discrete ¡collec8on ¡of ¡support ¡ points, ¡up ¡to ¡1 ¡point ¡per ¡subject. ¡

  • Each ¡NP ¡model ¡parameter ¡support ¡point ¡

contains ¡an ¡es8mate ¡of ¡each ¡model ¡ parameter ¡value, ¡and ¡of ¡the ¡probability ¡of ¡ that ¡collec8on ¡of ¡values ¡in ¡the ¡popula8on. ¡

  • No ¡assump8ons ¡need ¡to ¡be ¡made ¡about ¡the ¡

shape ¡of ¡the ¡distribu8ons ¡in ¡the ¡popula8on. ¡ They ¡use ¡UNCONSTRAINED ¡parameter ¡ distribu8ons ¡(UPD). ¡

slide-5
SLIDE 5

What ¡is ¡the ¡IDEAL ¡Popula8on ¡Model? ¡

  • The ¡correct ¡structural ¡PK/PD ¡Model. ¡
  • The ¡collec8on ¡of ¡each ¡subject’s ¡exactly ¡

known ¡parameter ¡values ¡for ¡that ¡model. ¡

  • Therefore, ¡mul8ple ¡discrete ¡individual ¡

models, ¡one ¡model ¡from ¡each ¡subject. ¡

  • Usual ¡con8nuous ¡sta8s8cal ¡summaries ¡can ¡

also ¡be ¡obtained, ¡but ¡usually ¡will ¡lose ¡info. ¡

  • NP ¡-­‑ ¡UPD ¡models ¡best ¡approach ¡this ¡

una_ainable ¡ideal. ¡

slide-6
SLIDE 6

An ¡NP-­‑UPD ¡Popula8on ¡ ¡Model, ¡made ¡by ¡Mallet ¡

slide-7
SLIDE 7

¡ ¡ ¡ ¡ ¡ ¡ ¡ ¡ ¡Tool ¡# ¡2 - Multiple Model (MM) dosage design ¡

Select ¡your ¡pa8ent’s ¡specific ¡therapeu8c ¡target ¡goal ¡– ¡not ¡ a ¡window. ¡ Use ¡all ¡mul8ple ¡models ¡in ¡NP ¡pop ¡model ¡(the ¡blue ¡slide). ¡ Give ¡a ¡candidate ¡dosage ¡regimen ¡to ¡each ¡model. ¡ Predict ¡future ¡results ¡from ¡each ¡model, ¡each ¡predic8on ¡ weighted ¡by ¡its ¡probability. ¡ Compute ¡WEIGHTED ¡SQUARED ¡ERROR ¡of ¡the ¡predic8ons ¡

  • verall ¡failure ¡to ¡hit ¡the ¡target. ¡

Find ¡the ¡regimen ¡having ¡LEAST ¡weighted ¡squared ¡error ¡in ¡ target ¡goal ¡achievement. ¡The ¡most ¡precise ¡regimen. ¡ It ¡uses ¡the ¡en8re ¡distribu8on ¡of ¡each ¡parameter, ¡not ¡just ¡a ¡ single ¡value ¡summary ¡of ¡some ¡assumed ¡distribu8on. ¡

slide-8
SLIDE 8

5/1/15 ¡ 8 ¡

Con8nuous ¡IV ¡Vanco. ¡Predic8ons ¡when ¡regimen ¡based ¡on ¡ CPD ¡parameter ¡means ¡is ¡given ¡to ¡all ¡support ¡points ¡

slide-9
SLIDE 9

5/1/15 ¡ 9 ¡

Vanco, ¡con8nuous ¡IV. ¡Predic8ons ¡when ¡MM ¡ regimen ¡is ¡given ¡to ¡all ¡support ¡points ¡

slide-10
SLIDE 10

No8ce ¡that… ¡

¡ ¡The ¡dose ¡itself ¡is ¡a ¡most ¡important ¡tool ¡to ¡

minimize ¡pa8ent ¡variability ¡in ¡response. ¡ ¡ ¡This ¡is ¡at ¡least ¡as ¡important ¡as ¡having ¡addi8onal ¡ covariate ¡informa8on. ¡ ¡ ¡ ¡With ¡parametric ¡models ¡you ¡will ¡never ¡be ¡aware ¡

  • f ¡this ¡issue, ¡and ¡never ¡be ¡able ¡to ¡do ¡it. ¡

¡

Jelliffe ¡R: ¡Goal-­‑Oriented, ¡Model-­‑Based ¡Drug ¡Regimens: ¡SePng ¡Individualized ¡Goals ¡for ¡ each ¡PaTent. ¡Therap. ¡Drug ¡Monit. ¡22: ¡325-­‑329, ¡2000. ¡ Jelliffe ¡R, ¡Bayard ¡D, ¡Milman ¡M, ¡Van ¡Guilder ¡M, ¡and ¡Schumitzky ¡A: ¡Achieving ¡Target ¡ Goals ¡most ¡Precisely ¡using ¡Nonparametric ¡Compartmental ¡Models ¡and ¡"MulTple ¡ Model" ¡Design ¡of ¡Dosage ¡Regimens. ¡Therap. ¡Drug ¡Monit. ¡ ¡22: ¡346-­‑353, ¡2000. ¡ Bustad ¡A, ¡Terziivanov ¡D, ¡Leary ¡R, ¡Port ¡R, ¡Schumitzky ¡A, ¡and ¡Jelliffe ¡R: ¡Parametric ¡and ¡ Nonparametric ¡PopulaTon ¡Methods: ¡Their ¡ComparaTve ¡Performance ¡in ¡Analysing ¡ a ¡Clinical ¡Data ¡Set ¡and ¡Two ¡Monte ¡Carlo ¡SimulaTon ¡Studies. ¡Clin. ¡Pharmacokinet., ¡ 45: ¡365-­‑383, ¡2006. ¡

slide-11
SLIDE 11

Tool # 3 - Correct description of assay errors

Describe ¡Assay ¡Errors ¡by ¡1/variance, ¡NOT ¡CV%. ¡

  • Get ¡es8mate ¡of ¡SD ¡for ¡every ¡TDM ¡serum ¡level ¡

going ¡through ¡the ¡lab ¡assay ¡system. ¡

  • Variance ¡= ¡SD2 ¡
  • 1/Variance ¡= ¡correct ¡Weight. ¡
  • Assay ¡Error ¡Polynomial ¡(AEP): ¡

¡ ¡ ¡SD ¡= ¡A0C0 ¡+ ¡A1C1 ¡+ ¡A2C2 ¡+ ¡A3C3 ¡

  • Store ¡it ¡in ¡the ¡solware. ¡
  • No ¡need ¡to ¡censor ¡low ¡data ¡any ¡more ¡
  • No ¡LLOQ! ¡This ¡is ¡an ¡illusion ¡from ¡using ¡CV%. ¡

Jelliffe ¡RW, ¡Schumitzky ¡A, ¡Van ¡Guilder ¡M, ¡Liu ¡M, ¡Hu ¡L, ¡Maire ¡P, ¡Gomis ¡P, ¡Barbaut ¡X, ¡ and ¡Tahani ¡B: ¡Individualizing ¡Drug ¡Dosage ¡Regimens: ¡Roles ¡of ¡PopulaTon ¡ PharmacokineTc ¡and ¡Dynamic ¡Models, ¡Bayesian ¡FiPng, ¡and ¡AdapTve ¡Control. ¡ TherapeuTc ¡Drug ¡Monitoring, ¡15: ¡380-­‑393, ¡1993. ¡

slide-12
SLIDE 12

21.15 10.33 9.9663 14.323 0.56708 0.423024 0.413376 0.7973 1.7188

0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 2 2 4 8 12 Concentration (ug/ml) SD 5 10 15 20 25 CV%

CV% SD

slide-13
SLIDE 13

Tool # 4 – Estimate CCr from changing SCr ¡

0.4W(C2 ¡– ¡C1) ¡= ¡P ¡– ¡Cavg ¡x ¡CCr ¡x ¡1440 ¡ (Daily ¡change ¡= ¡Produc8on ¡– ¡Excre8on) ¡

  • Where ¡

P ¡is ¡daily ¡crea8nine ¡produc8on ¡(see ¡ar8cle) ¡ C1 ¡and ¡C2 ¡are ¡SCr ¡in ¡mg/100ml, ¡ W ¡is ¡weight ¡(in ¡hundreds ¡of ¡grams), ¡and ¡ CCr ¡is ¡crea8nine ¡clearance ¡(in ¡hundreds ¡of ¡ ¡ml/ min). ¡ Then ¡adjust ¡for ¡1.73 ¡M2 ¡BSA ¡ ¡

Jelliffe ¡R: ¡Es8ma8on ¡of ¡Crea8nine ¡Clearance ¡in ¡Pa8ents ¡with ¡Unstable ¡ Renal ¡Func8on, ¡without ¡a ¡Urine ¡Specimen. ¡Am. ¡J. ¡Nephrology, ¡22: ¡ 3200-­‑3204, ¡2002. ¡

¡ ¡

slide-14
SLIDE 14

Es8ma8on ¡errors ¡in ¡CCr ¡

CV ¡of ¡serum ¡crea8nine ¡= ¡± ¡5% ¡ CV ¡of ¡urine ¡crea8nine ¡= ¡± ¡8% ¡ Then ¡if ¡collect ¡24 ¡hr ¡urine ¡with ¡CV ¡= ¡± ¡5%, ¡ 52=25 ¡twice ¡= ¡50, ¡plus ¡82=64, ¡= ¡114. ¡ SQRT ¡114 ¡~ ¡11 ¡ So ¡the ¡classical ¡CCr ¡has ¡a ¡CV ¡= ¡± ¡11%, ¡ ¡

  • r ¡95% ¡conf ¡limits ¡of ¡± ¡22% ¡
slide-15
SLIDE 15

Es8mated ¡vs ¡Measured ¡CCr ¡

slide-16
SLIDE 16

Tool # 5 - MM Optimal TDM protocols ¡

Do ¡NOT ¡wait, ¡and ¡do ¡not ¡get ¡only ¡steady ¡state ¡trough ¡levels! ¡ Start ¡right ¡with ¡the ¡very ¡first ¡dose! ¡ Consider ¡1 ¡sample ¡for ¡each ¡model ¡parameter ¡to ¡be ¡es8mated. ¡ Use, ¡for ¡example, ¡D-­‑op8mal ¡design ¡strategies. ¡There ¡is ¡an ¡

  • p8mal ¡8me ¡to ¡sample ¡for ¡each ¡model ¡parameter ¡value, ¡

given ¡a ¡certain ¡dosage ¡regimen ¡format. ¡ Olen, ¡get ¡a ¡peak ¡and ¡a ¡sample ¡at ¡about ¡1/3 ¡the ¡peak. ¡ Be_er, ¡consider ¡the ¡new ¡MM-­‑op8mal ¡design, ¡op8mizing ¡the ¡ 8mes ¡to ¡maximize ¡precision ¡of ¡dosing ¡to ¡desired ¡target ¡ conc, ¡or ¡AUC, ¡for ¡example. ¡ Do ¡NOT ¡waste ¡money, ¡effort, ¡and ¡pa8ent ¡care ¡with ¡poor ¡

  • designs. ¡The ¡TDM ¡community ¡can ¡do ¡a ¡MUCH ¡BETTER ¡job ¡
  • here. ¡ ¡
slide-17
SLIDE 17

5/1/15 USC LAPK 17

Best ¡8me ¡ ¡ to ¡sample ¡ For ¡a ¡constant ¡assay ¡error, ¡the ¡greatest ¡change ¡in ¡serum ¡conc ¡when ¡volume ¡of ¡ dist ¡changes ¡is ¡at ¡the ¡true ¡peak, ¡not ¡later. ¡Samples ¡near ¡the ¡peak ¡are ¡most ¡ informa8ve ¡about ¡Vd. ¡

Example ¡of ¡D-­‑OpTmal ¡Design: ¡ ¡ Finding ¡Volume ¡Vd ¡from ¡a ¡First-­‑Order ¡Response ¡

slide-18
SLIDE 18

5/1/15 USC LAPK 18

The ¡greatest ¡change ¡in ¡serum ¡conc ¡from ¡a ¡change ¡in ¡the ¡Kel ¡is ¡when ¡the ¡serum ¡ conc ¡is ¡36% ¡of ¡the ¡original ¡peak ¡value. ¡Samples ¡near ¡1.44 ¡half-­‑8mes ¡are ¡most ¡ informa8ve ¡about ¡Kel. ¡

Example ¡of ¡D-­‑OpTmal ¡Design: ¡ ¡ Finding ¡Kel ¡from ¡First-­‑Order ¡Response ¡

Best ¡8me ¡ ¡ to ¡sample ¡

slide-19
SLIDE 19

Mul8ple ¡Model ¡Op8mal ¡Design ¡

– Optimizing experimental design this way is fundamentally different from

  • ptimizing experimental design to estimate patient’s parameters, as in

D-optimal design, etc.

0.02 0.04 0.06 0.08 0.1 0.12 1.5 2 2.5 0.005 0.01 0.015 0.02 Vs 10 Model Prior Kel Prob

MM Prior V K 1 - Randomly sample a population model support point. 2 - Give it a simulated dose and some simulated

  • bservations, at a specific set of candidate

sampling times, with assay noise. 3 - Get the MAP Bayesian posterior estimate. 4 – See if that MAP estimate corresponds to the

  • riginal point (correct classification) or some other

point (incorrect). 5 – Repeat steps 1-4 many times. See the percent

  • f times the classification is correct.

6 – Do this for many candidate designs. 7 - Find the sampling design that minimizes the percent incorrect classification.

slide-20
SLIDE 20

20

  • Model Response Separation r(t) is the

separation between two model responses at a given time t

  • Defines natural statistic for discriminating

between two models

  • Bayes Risk of mis-classification is

shown in gray area

Model Response Separation r(t)

2 4 6 8 10 0.2 0.4 0.6 0.8 1

Model Responses

Time Model 1 Reponse Model 2 Response

r(t) = response separation

  • 1
  • 0.5

0.5 1 1.5 0.5 1 1.5 2

Two-Model Classification Problem

r(t)=response separation

Bayes Risk

  • Bayes Risk (gray area)

decreases as response separation r(t) increases

  • Models are best

discriminated by sampling at a time t that maximizes r(t) Pull Gaussians apart to minimize gray area

slide-21
SLIDE 21

21 21

ED EID API MMOpt

Invariant under regular linear reparametrization*

Yes Yes Yes Yes

Invariant under regular nonlinear reparametrization*

No No Yes Yes

Allows taking fewer than p samples, p= # of parameters

No No No Yes

Can handle heterogeneous model structures

No No No Yes

Gives known optimal solution to 2-model example*

No No No Yes

Captures main elements of minimizing Bayes risk

No No No Yes

Comparison Table

*Proved in Bayard et. al., PODE 2013 [23]

slide-22
SLIDE 22

22

Weighted ¡MMOpt ¡

Optimal Dosing Parameter Estimation Metric Estimation

slide-23
SLIDE 23

23 23

Model Parameters # K V

1 0.090088 113.7451 2 0.111611 93.4326 3 0.066074 90.2832 4 0.108604 89.2334 5 0.103047 112.1093 6 0.033965 94.3847 7 0.100859 109.8633 8 0.023174 111.7920 9 0.087041 108.6670 10 0.095996 100.3418

PK Population Model with 10 Multiple Model Points - First-Order PK Model

5 10 15 20 25 0.5 1 1.5 2 2.5 3 3.5 All subject responses Time (hr) Concentration

5 10 15 20 50 100 150 200 250 300 350 400 Dose Input Time (hr) Units

Dose input = 300 units for 1 hour, starting at time 0 Model Responses

  • Grid points 15 min apart
  • MMOpt optimized over time grid
slide-24
SLIDE 24

24

Unweighted ¡MMOpt ¡for ¡PK ¡Es8ma8on ¡

  • Summary of optimal 1,2 and 3 sample designs applied to PK Estimation
  • 1 Sample Design: MMOpt performance equals Bayesian optimal design (both

have Bayes Risk of 0.5474).

  • MMOpt performance improves on EDopt design for 2 and 3 sample designs

– 2 Sample Design: Bayes Risk of 0.29 versus 0.33 – 3 Sample Design: Bayes Risk of 0.23 versus 0.26

  • All results are statistically significant to p<0.0001
slide-25
SLIDE 25

25

Weighted ¡MMOpt ¡for ¡AUC ¡Control ¡

slide-26
SLIDE 26

26

Weighted ¡MMOpt ¡for ¡AUC ¡Control ¡(Cont’d) ¡

  • Summary of optimal 1,2 and 3 sample designs applied to AUC control
  • 1 Sample Design: weighted MMOpt performance approximates that of the

weighted Bayesian optimal design (RMS error of 3.62 versus 3.77 AUC units)

  • MMOpt performance improves on EDopt design for 2 and 3 sample designs

– 2 Sample Design: RMS error of 2.11 versus 2.62 (units of AUC) – 3 Sample Design: RMS error of 1.70 versus 2.42 (units of AUC)

  • All results are statistically significant to p<0.0001
slide-27
SLIDE 27

Tool # 6 - NP Bayesian adaptive control tools

Previous ¡data ¡= ¡past ¡experience ¡= ¡popula8on ¡model. ¡ ¡ 1 ¡– ¡Pop ¡model ¡is ¡the ¡Bayesian ¡Prior ¡(prior ¡to ¡now) ¡– ¡ ¡es8mates ¡probability ¡of ¡model ¡parameters ¡based ¡on ¡past ¡experience. ¡ 2 ¡-­‑ ¡New ¡data ¡from ¡pa8ent. ¡Recompute ¡param ¡ ¡ ¡distribu8ons ¡based ¡on ¡past ¡+ ¡new. ¡ 3 ¡-­‑ ¡This ¡is ¡the ¡Bayesian ¡posterior ¡– ¡the ¡individualized ¡ ¡pt ¡model. ¡ ¡

slide-28
SLIDE 28

An ¡Ini8al ¡Gentamicin ¡Predic8on ¡based ¡on ¡ popula8on ¡model ¡prior: ¡goal ¡= ¡12 ¡ug/ml ¡

95 Percentile Distance

2 4 6 8 10 12 14 5 10 15 20 25 Probability [%] Concentration in central compartment [ug/mL]

slide-29
SLIDE 29

Gevng ¡Individualized ¡Bayesian ¡ Models ¡in ¡Pa8ents ¡– ¡4 ¡ways ¡

¡

  • 1. MM ¡Bayesian ¡posteriors ¡
  • 2. MAP ¡Bayesian ¡Posteriors ¡
  • 3. Hybrid ¡Bayesian ¡posteriors. ¡
  • 4. Interac8ng ¡MM ¡(IMM) ¡Sequen8al ¡

¡Bayesian ¡Posteriors. ¡

¡

¡ ¡ ¡

slide-30
SLIDE 30
slide-31
SLIDE 31

Es8mates ¡with ¡MM ¡posterior ¡

slide-32
SLIDE 32

Hybrid ¡Bayesian ¡posterior ¡

  • Start ¡with ¡MAP ¡Bayesian. ¡It ¡reaches ¡out, ¡but ¡

not ¡fully ¡(shrinkage). ¡Pop ¡prior ¡holds ¡it ¡back. ¡

  • Add ¡new ¡support ¡points ¡nearby, ¡inside ¡and ¡

mostly ¡outside, ¡to ¡AUGMENT ¡the ¡pop ¡model ¡ for ¡the ¡coming ¡new ¡pa8ent ¡data. ¡

  • Then ¡do ¡MM ¡Bayesian ¡on ¡ALL ¡the ¡support ¡
  • points. ¡
  • Can ¡also ¡use ¡for ¡pa8ents ¡without ¡a ¡

popula8on ¡model. ¡ ¡

slide-33
SLIDE 33

Pa8ent ¡on ¡digoxin ¡and ¡quinidine ¡-­‑ ¡ Hybrid ¡Bayesian ¡posterior ¡plot ¡

slide-34
SLIDE 34

Pa8ent ¡on ¡digoxin ¡and ¡quinidine ¡-­‑ ¡Hybrid ¡ Bayesian ¡Posterior ¡Joint ¡Density ¡

slide-35
SLIDE 35

Interac8ng ¡Mul8ple ¡Model ¡(IMM) ¡ Sequen8al ¡Bayesian ¡analysis. ¡

  • All ¡other ¡fivng ¡methods ¡assume ¡fixed ¡parameter ¡

values ¡throughout ¡the ¡data ¡analysis. ¡

  • IMM ¡lets ¡param ¡distribu8ons ¡change ¡with ¡each ¡new ¡

data ¡point ¡to ¡get ¡sequen8al ¡Bayesian ¡posterior ¡

  • distribu8ons. ¡ ¡
  • IMM ¡tracks ¡Gent ¡and ¡Vanco ¡behavior ¡best ¡in ¡ICU ¡
  • pa8ents. ¡
  • Bayard ¡D, ¡and ¡Jelliffe ¡R: ¡A ¡Bayesian ¡Approach ¡to ¡Tracking ¡PaTents ¡having ¡Changing ¡

PharmacokineTc ¡Parameters. ¡J. ¡Pharmacokin. ¡Pharmacodyn. ¡31 ¡(1): ¡75-­‑107, ¡2004. ¡

  • Macdonald ¡I, ¡ ¡Staatz ¡C, ¡Jelliffe ¡R, ¡and ¡Thomson ¡A: ¡EvaluaTon ¡and ¡Comparison ¡of ¡Simple ¡

MulTple ¡Model, ¡Richer ¡Data ¡MulTple ¡Model, ¡and ¡SequenTal ¡InteracTng ¡MulTple ¡Model ¡ (IMM) ¡Bayesian ¡Analyses ¡of ¡Gentamicin ¡and ¡Vancomycin ¡Data ¡Collected ¡From ¡PaTents ¡ Undergoing ¡Cardiothoracic ¡Surgery. ¡Ther. ¡Drug ¡Monit. ¡30:67–74, ¡2008. ¡

slide-36
SLIDE 36
slide-37
SLIDE 37
slide-38
SLIDE 38
slide-39
SLIDE 39

By ¡avoiding ¡clearance ¡and ¡using ¡Vs ¡and ¡Ks ¡instead, ¡the ¡2 ¡separate ¡issues ¡

  • f ¡fluid ¡balance ¡and ¡drug ¡eliminaTon ¡can ¡be ¡managed ¡best. ¡
slide-40
SLIDE 40

TDM ¡up ¡to ¡now ¡has ¡been ¡mostly ¡for ¡ achieving ¡general ¡serum ¡concentra8on ¡

  • guidelines. ¡
  • But ¡now, ¡what ¡about ¡seeing ¡each ¡pa8ent ¡as ¡a ¡

unique ¡individual? ¡

  • Pa8ents ¡vary ¡greatly ¡in ¡their ¡sensi8vity ¡to ¡

serum ¡concentra8ons. ¡

  • Also, ¡digoxin ¡clinical ¡effect ¡does ¡not ¡correlate ¡

with ¡serum ¡concentra8ons, ¡but ¡with ¡ peripheral, ¡nonserum ¡compartment, ¡

  • concentra8ons. ¡
slide-41
SLIDE 41

Serum digoxin concentrations in nontoxic and toxic patients found by Doherty [1]. Note great overlap between therapeutic and toxic concentrations, and the fact that approximately half the patients with serum levels of 3.0 ng/ ml or more were NOT toxic. Also note that the incidence of toxicity is very low for levels up to 1.0 ng/ml, moderate (though significant) for levels of 1.0 to 2.0, more above 2.0, but still only about 50% for levels of 3.0 ng/ml or greater.

slide-42
SLIDE 42

Management ¡of ¡digoxin ¡therapy ¡

  • Two ¡compartment ¡NP ¡pop ¡model. ¡
  • Clinical ¡effect ¡correlates ¡with ¡peripheral ¡

compartment ¡concentra8ons, ¡not ¡with ¡central ¡ (serum) ¡concentra8ons. ¡

  • Plan ¡regimens ¡either ¡for ¡trough ¡serum ¡

concentra8ons ¡(usually ¡0.7 ¡ng/ml) ¡or ¡peak ¡ peripheral ¡concentra8ons ¡(usually ¡7 ¡ng/gm, ¡7 ¡ hrs ¡aler ¡an ¡oral ¡dose). ¡

  • But ¡pa8ents ¡with ¡aur ¡fib, ¡flu_er ¡olen ¡need ¡9-­‑18 ¡

ng/gm ¡to ¡convert ¡and ¡stay ¡converted. ¡

slide-43
SLIDE 43

Bauman ¡et ¡al: ¡Am.J.Cardiovascular ¡Drugs ¡2006; ¡6: ¡

slide-44
SLIDE 44

Literature ¡data ¡

  • Study

¡ ¡ ¡ ¡ ¡ ¡ ¡ ¡ ¡ ¡Digoxin ¡/Placebo ¡ ¡ ¡ ¡ ¡Periph ¡conc. ¡ ¡ ¡ ¡ ¡P ¡

  • Falk ¡et ¡al ¡

¡ ¡ ¡ ¡ ¡9/18 ¡v ¡8/18 ¡ ¡8.0 ¡ ¡ ¡ ¡ ¡ ¡ ¡ ¡ ¡NS ¡

  • DAAF

¡ ¡ ¡ ¡ ¡ ¡ ¡ ¡ ¡ ¡ ¡ ¡ ¡60/117 ¡v ¡56/122 ¡ ¡ ¡ ¡ ¡ ¡ ¡ ¡ ¡ ¡9.23 ¡ ¡ ¡ ¡ ¡ ¡ ¡ ¡ ¡NS ¡

  • Jordaens ¡et ¡al ¡ ¡ ¡ ¡ ¡ ¡9/19 ¡v ¡8/20 ¡ ¡ ¡ ¡ ¡ ¡ ¡ ¡ ¡ ¡ ¡ ¡11.25 ¡ ¡ ¡ ¡ ¡ ¡ ¡ ¡ ¡ ¡ ¡NS ¡

¡ ¡ ¡ ¡ ¡ ¡ ¡ ¡ ¡ ¡ ¡ ¡ ¡Conversion ¡to ¡RSR ¡

  • Hou ¡et ¡al ¡ ¡ ¡ ¡ ¡ ¡ ¡ ¡ ¡ ¡ ¡ ¡ ¡ ¡ ¡ ¡ ¡17/24 ¡(71%) ¡ ¡ ¡ ¡ ¡ ¡ ¡ ¡ ¡ ¡ ¡ ¡ ¡ ¡ ¡9.0 ¡ ¡ ¡ ¡ ¡ ¡ ¡ ¡ ¡<0.05 ¡
  • Weiner ¡et ¡al

¡ ¡ ¡ ¡ ¡ ¡40/47 ¡(85%) ¡ ¡ ¡ ¡ ¡ ¡ ¡ ¡ ¡ ¡ ¡13.9 ¡ ¡ ¡<<0.005 ¡

¡

(Clinical ¡PharmacokineTcs ¡(DOI) ¡10.1007/s40262-­‑014-­‑0141-­‑6) ¡

slide-45
SLIDE 45

New ¡onset ¡AF ¡– ¡ ¡SCr ¡= ¡0.8, ¡ ¡ Serum ¡concentra8ons ¡over ¡8me ¡ ¡

slide-46
SLIDE 46

New ¡onset ¡AF ¡– ¡ ¡SCr ¡= ¡0.8, ¡ ¡ Peripheral ¡concentra8ons ¡over ¡8me ¡

slide-47
SLIDE 47

New ¡onset ¡AF ¡– ¡96 ¡F ¡ ¡SCr ¡= ¡8, ¡CCr ¡= ¡3.7 ¡ Serum ¡concentra8ons ¡over ¡8me ¡

0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 50 10 15 20 25 30 35 Concentration in central compartment [ug/mL] Tim e [h

  • u

rs] W gtA v g

slide-48
SLIDE 48

New ¡onset ¡AF ¡– ¡96 ¡F ¡ ¡SCr ¡= ¡8, ¡CCr ¡= ¡3.7 ¡ Peripheral ¡concentra8ons ¡over ¡8me ¡

2 4 6 8 10 12 50 10 15 20 25 30 35 Concentration in peripheral compartment [mg/kg] Tim e [h

  • u

rs] W gtA v g

slide-49
SLIDE 49

Problems ¡with ¡educa8on ¡in ¡drug ¡therapy ¡

  • No ¡one ¡teaches ¡meaningful ¡PK ¡in ¡med ¡schools ¡

anywhere ¡

  • No ¡one ¡knows ¡anything ¡about ¡dosage ¡

individualiza8on ¡

  • Their ¡only ¡informa8on ¡comes ¡from ¡the ¡drug ¡

companies ¡

  • Poor ¡illiterate ¡med ¡schools! ¡
  • Poor ¡illiterate ¡docs! ¡
  • Poor ¡poorly ¡treated ¡pa8ents! ¡
  • Rich ¡and ¡richer ¡drug ¡companies! ¡
  • Rich ¡and ¡richer ¡lawyers! ¡
slide-50
SLIDE 50

¡ ¡ ¡ ¡ ¡ ¡ ¡ ¡ ¡Problems ¡with ¡drug ¡marke8ng ¡

slide-51
SLIDE 51

Suppressing ¡their ¡own ¡science ¡for ¡ marke8ng! ¡

  • They ¡have ¡the ¡data ¡for ¡each ¡pa8ent ¡-­‑ ¡
  • Outcome, ¡and ¡An8platelet ¡effect. ¡
  • They ¡can ¡easily ¡find ¡the ¡an8platelet ¡effect ¡

associated ¡with ¡best ¡outcome, ¡least ¡bleeding. ¡

  • They ¡can ¡set ¡a ¡target ¡an8platelet ¡effect, ¡
  • And ¡dose ¡each ¡pa8ent ¡to ¡achieve ¡it! ¡
  • Why ¡do ¡they ¡adver8se ¡that ¡this ¡is ¡NOT ¡

NEEDED?? ¡Why?? ¡Think ¡about ¡it! ¡

  • WE ¡BADLY ¡NEED ¡INDIVIDUALIZED ¡DOSAGE. ¡