Transcript Drug Dosing in PCRRT - Pediatric Continuous Renal Replacement

Drug Dosing in
PCRRT
Deb Pasko, Pharm.D
Pharmacy Clinical Specialist, PICU
University of Michigan Health System
CRRT Solute Removal

Lots of things
removed by CRRT!

Drugs, nutrients…
FD&C Blue dye
#1…

Crit Care Med, Mar
Diffusive Therapies


Dialysate is used (lactated Ringers, PD
solution, etc)
Good for small solute removal (<500 Da)


diffusion rate inversely proportional to MW
Efficiency of solute removal dependent on
Blood flow
 Dialysate flow
 Filter type
 Solute molecular weight


Less good for larger solutes (MM,
Vancomycin?)
Joy MS, Matzke GR, Frye
RF, Palevsky PM. AJKD
1998;31:1019-27.
RRT Drug Removal Mechanisms



Diffusion
Convection
Adsorption

May be important for 2 Microglobulin removal


Especially for PMMA membranes
Rarely important for drugs
Vancomycin overdose

16 day old full-term infant presented to OSH
hypothermia, bradycardia, and hypovolemia.
Progressed to develop cardiac arrest, transferred
to U of M. Dry wt. 2kg.

Received 3 doses of vancomycin 100mg/kg
Initial vancomycin serum concentration was

195.5 mg/L (desired peak conc ~35mg/L)
Vancomycin overdose case
Hemodialyzer differences: Are they
important in CVVHD?

Most published drug dosing guidelines assume
they are all equal in terms of drug removal
most hemofilters are high-permeability with large
pores
 frequently high flux hemodialyzers were used for
CRRT


Vancomycin CVVHD clearance differences
between different hemodialyzers

Joy MS, et al. Am J Kidney Dis 1998 Jun;31(6):1019-27
Convective Therapies
(Hemofiltration)


No dialysate, removes plasma water as it seeps
through membrane
Removes small and large molecules easily




as long as they can fit through membrane
Protein binding important determinant – sieving coefficient
<15,000 Da has potential to be removed substantially
Drug removal easy to calculate


based on sieving coefficient – usually a function of PPB
ultrafiltrate concentration/plasma concentration
Doses Derived Via Sieving
Coefficient



Sieving Coefficient (SC) known for many drugs
 SC= UF/A
Comes from CAVH or CVVH data
Assumption often made that SC can be used
CVVHD when dialysate rate is low.


Saturation coefficient (Sa) more properly used in
CVVHD
SC related to protein binding of drugs

Protein binding may differ in critically ill vs. normals
Sieving Coefficient & Protein
Binding
Drug
Amikacin
Imipenem
Metronidazole
Penicillin
Ranitidine
Vancomycin
Valproic Acid
Reported SC
Free Fraction
0.93
0.78
0.84
0.68
0.78
0.80
0.22
0.95
0.80
0.80
0.50
0.85
0.90
0.10
Drug Dosing recommendations
based on Sieving Coefficient (SC)

Clearance total = ClCRRT + Cl residual renal + Cl nonrenal


SC equations only account for ClCRRT
What about other clearances?
 Cl residual renal usually not an issue in CRRT
patients
 Cl non-rena l not always available for drugs
Non-renal clearance rates of
selected drugs in patients with
normal renal function and ESRD
NORMAL RENAL FX
(ML/MIN/70 KG)
ESRD
% DECLINE IN CL
Acyclovir
65
29
55
Aztreonam
40
27
33
Cefotaxime
217
130
40
Imipenem
128
54
56
Procainamide
257
102
60
Vancomycin
40
6
85
DRUG
CRRT Challenges: Drug Dosing

Does CVVH removal = CVVHDF = CVVHD???
 Molecular weight determines whether solute diffuses well
 Vancomycin (MW 1450 Da)
 Aminoglycosides (MW ~450 Da)
 High dialysate flow rates don’t allow sufficient time for
diffusion
 Probably not an issue when flow = ~1000ml/1.73m2/hr

Does Sieving Coefficient (CVVH) = Saturation
Coefficient (CVVHD)???
CRRT Challenges: Drug Dosing





NO PEDIATRIC DOSING!!!!!!!
Most CRRT dosing guidelines based on CVVH
@ UFR of 1000 mL/hr
Trend is for higher UFR and HD flows
UM uses 2L/1.73m2/hr
Higher flow rates now achievable with new
machines
 solute
removal (H, HD, HDF) mechanisms
Pediatric CrCl

CLCR = K x L/SCR
Where ClCR = creatinine clearance in
ml/min/1.73m2
K = constant of proportionality age specific
Age
K
LBW ≤ 1yo
0.33
Full-term ≤ 1yo
0.45
2-12
0.55
13-21 female
0.55
13-21 male
0.70

Calculating Total Clearance

Example:
2yo, 15kg, L = 60cm, SCR = 1.0 mg/dL, K = 0.55
 CrCl = 0.55 x 60/1 = 33ml/min/1.73m2
 However, if anuric renal clearance = zero
 PCRRT = CVVHD of 2L/1.73m2/hr, BSA of 0.5


Qd = ~578ml/hr


(38.5ml/kg, or 9.6ml/min/0.5 BSA, or 33.2ml/min/1.73ml/min)
If this patient was not anuric and had renal fxn as above =
66ml/min/1.73m2, and we need to adjust accordingly
Adjusting doses based on Cl

Using the previous example for vancomycin:
>50ml/min/1.73m2 = q6-8h dosing
 30-50ml/min/1.73m2 = q12h dosing

It is easy to under-dose or possibly overdose using
this method, need to be careful
Is CrCl the most reliable method for children?
What drugs do we care about?

Drugs are “dialyzable” if:
 Small MW
 Small volume of distribution
 Not highly protein bound
 Water soluble
Case





10mof, ALL s/p chemo & BMT x60days, now admitted to
unit for increased O2 needs requiring vent support, GVHD
gut/liver stage IV and in septic shock.
PE: T 39.1, HR 180, BP 60/30, wt. 8.5kg, Ht 60cm
I/O: 900/50 over past 24hrs (0.24cc/kg/hr)
Baseline Scr = 0.3mg/dL, now 0.6mg/dL
Meds: Dopamine, Cefepime, Gentamicin, Linezolid,
Voriconazole, Pentamidine, Hydrocortisone, Protonix,
TPN/lipids, Dilaudid/Ativan, Phenobarb
Case con’t


AM BC shows Pseudomonas aer. and VRE
Order written to start CVVH @ 2L/1.73m2/hr,
Calc. clearance: BSA = 0.38m2 (7.24ml/min, or
33ml/min/1.73m2)
 What drugs do we care about?
 If you can titrate we don’t necessarily care
 For this patient antibiotics are going to save her
life

So what drugs need adjustment?






Dopamine?
Cefepime?
Gentamicin?
Linezolid?
Voriconazole?
Pentamidine?





Hydrocortsione?
Protonix?
TPN?
Dilaudid?
Ativan?
Antibiotic Guidelines UM
Linezolid Clearance During CVVHDF




85 yo 90 kg anuric male in the SICU with
documented abdominal VRE infection
Linezolid 600 mg IV q12
No published literature on CRRT removal
CVVHDF regimen:
dialysate flow rate 2000 mL/hr
 mean ultrafiltrate production rate of 775 mL/hr

Linezolid Calculations
Half-life, elimination rate, and volume of distribution
 Sieving coefficient (SC) was calculated:
SC= CE / Cp, Cp = (CA + CV) / 2

The clearance from CRRT (Cl CRRT) calculated as:
Cl CRRT = (QD + QF) x SC




CE = the concentration in the effluent
Cp is the linezolid concentration in the plasma
CA is the linezolid concentration in the plasma drawn from
the pre-filter sampling port
CV is the linezolid concentration in the plasma drawn from
the post-filter sampling port.
Linezolid Results




Vd = 60L (normal 40-60L)
T1/2 = 7.5-9 hrs during CVVHDF (8hrs)
SC = 0.77– 0.81 (PPB 30%)
ClCRRT = 36.5 mL/min with mean effluent flow
rate of 46.2 mL/min




(normal ClR 40mL/min)
No dosage change necessary
First measured linezolid CRRT report
Kraft MK, Pasko DA, DePestel DD, Ellis JJ, Peloquin CA,
Mueller BA. Linezolid clearance during continuous venovenous
hemodiafiltration: A case report. Pharmacotherapy. 2003
Aug;23(8):1071-5.
So what drugs need adjustment?






Dopamine?
Cefepime?
Gentamicin?
Linezolid?
Voriconazole?
Pentamidine?





Hydrocortsione?
Protonix?
TPN?
Dilaudid?
Ativan?
Gentamicin pharmacokinetics




This patient weighing 8.5kg receives a gent
dose of 21mg (2.5mg/kg)
What peak concentration (mg/L) can be
expected?
Volume of distribution of gent is 0.2-0.4L/kg
0.25L/kg is normal, but in fluid overloaded
patients, expect higher values. If 0.3L/kg =
2.55 Liters = Vd
 21mg/2.55L = 8.2 mg/L assuming no drug removal

Gent kinetics con’t


30 min after the 21mg dose is done a peak is
done = 4.0mg/L
What is the patients actual volume of
distribution?

5.1 Liters = 0.6L/kg (actually double!!!!)
Gent kinetics con’t


Peak was 4.0 mg/L
12 hours later a random level was done
1.0 mg/L
 What is the half-life (t1/2) of gentamicin?
 4.0mg/L → 2.0mg/L → 1.0mg/L in 12 hours
 6 hour half-life
 Ln 4 – ln 1 = kel = 0.115


12hrs
Gent kinetics FINAL

Half-life = 0.693 / 0.115 = 6 hours
So what drugs need adjustment?






Dopamine?
Cefepime?
Gentamicin?
Linezolid?
Voriconazole?
Pentamidine?





Hydrocortsione?
Protonix?
TPN?
Dilaudid?
Ativan?
Phenobarbital case







2 wof transferred to UM w/ severe CHF w/ AV valve
regurgitation and seizure dx
1/20 had cleft AV valve repair w/ PDA ligation, went
on VA ECMO, developed ARF
1/24 went on CVVHD in-line w/ECMO circuit
Wt 3.45kg (dry), Ht 47cm, BSA 0.21m2
Qd set at 300ml/hr (~2400ml/1.73m2/hr)
Quf at 69ml/hr (drips + no net loss)
Hemodiafilter = Mini-Plus, 0.08m2
Phenobarbital case


Phenobarbital dose pre dialysis initiation =
25mg q24h = 7.2mg/kg = 35.5mg/L serum
concentration
CVVHD started 1/24
1/25 Pb = 14.2 mg/L
 1/26 Pb = 9.6
 1/28 Pb = 13.9 @ 0700
 1/28 1400 sequential levels done

Phenobarbital case
Site
Concentration (mg/L)
Pre-oxygenator
26.5
Post-oxygenator
24.7
Pre-hemodiafilter
26.2
Post-hemodiafilter
25.9
Effluent
11.4
Drug dosing problems in
high volume hemofiltration

Most drugs have > 2 compartments (pools)
like urea measurements during HD
 high volume hemofiltration removes drug from
peripheral compartment rapidly
 how fast can drug transfer from deeper
compartment?


Many drugs rapidly stripped from first pool
Phenobarb case final


SC = 0.44
ClCRRT =
2.7ml/0.21m2/min or 22.3ml/1.73m2/min (40)
 Vd = 3.24L/kg (0.9L/kg, normal 0.6)
 New maintenance dose to maintain level of
25mg/L =

32.4mg (~10mg/kg) IV q8hrs
 Original maintenance dose 25mg q24hrs

Solute Kinetics during CRRT
Dialysate
UF Soln
UF Soln
Deeper
compartment(s)?
Pool 3
???Liters
Qb
Dialysate
Spent
Dialysate
&/or
UF
CRRT
Qb
Patient
k23
Peripheral
compartment
Deep/Central
Compartment
Pool 2
???Liters
Pool 1
6L?
k12
k32
k21
Mueller BA, Pasko DA. Artif Organs 2003;27:808-14.
IV drug administration: Drug
removed as it is infused
 Drug
is infused into compartment being
filtered/dialyzed
 reduced
ability to distribute into tissues (k12)
 serum concentrations during infusion higher than
usual “therapeutic” serum concentration
 6L/hr
= 1L/10 min = entire plasma
volume/hr
 Qb 150 ml/min Quf/hd 33 mL/min =
 22%
of volume removed
 “first-pass” effect
Drug Prescribing in Renal
Failure
edited by George Aronoff et al






Commonly carried text by
pharmacists
http://www.kdpbaptist.louisville.edu/renalbook/
New edition to come out soon
Recommendations for new drugs
IHD and CRRT recommendations
Pediatric recommendations
Strength of Evidence
A.
B.
C.
D.
Controlled studies in humans or
large case series experience
Small Case Series or Human 58 drugs
are A->C
Uncontrolled Trials
Animal or In Vitro Data
Known Drug Characteristics
A.
Vast majority are of this type
D. “Known drug characteristics“
 These
recommendations made by
panel of nephrologists and
pharmacists
 Based on:
 Protein
Binding Information
 Volume of Distribution
 Molecular Weight
When in doubt, start here…


Blood flow, filter type are not very important.
Find out

In CVVHD: Dialysate flow rate (ml/hr)


In CVVH: Substitution Fluid rate (ml/hr)




Usually 2 L/1.73m2/hr (33 mL/1.73m2/min)
Usually 2L/1.73m2/hr (33 mL/1.73m2/min)
Add this to patient’s native Cr Cl
(ml/1.73m2/min)
This is patient’s new Cr Cl  dose accordingly
Works in most cases…is good enough for initial
estimates. Follow up with drug level monitoring.
PCRRT & Roller pump
Future research needed
ECMO/PCRRT
 MARS
 RAD

CRRT Dosing should not be
confusing!