Transcript AKI and Biomarkers - Pediatric Continuous Renal Replacement

Pediatric Acute Kidney Injury and Biomarkers
Stuart L. Goldstein, MD
Professor of Pediatrics
Baylor College of Medicine
6th PCRRT Conference, Rome 2010
Acknowledgements
Baylor College of Medicine/Texas Children’s AKI Study Group
Laura Loftis, MD
Annabelle Chua, MD
Ayse Arikan, MD
Michael Zappitelli, MD
Alyssa Riley-Kothari, MD
Yue Du, PhD
Leticia Castillo, MD
Jack Price, MD
David Nelson, MD
John Lynn Jeffries, MD
Joshua Blinder, MD
Jeffrey Towbin, MD
Anjan Shah, MD
Brady Moffett, RPh
Outline
 AKI Epidemiology – Old Definitions
 Risk Factor Assessment
 New AKI Definitions
 Treatment
 Prognosis
 New Advancements - Biomarkers
AKI Definitions to 2002
 Over 30 definitions in published literature
 Nearly all based on absolute or change in serum creatinine
concentration
 Pediatric AKI definitions – All AKI is created equal
 100% rise in SCr
 eCCL < 75 ml/min/1.73m2
 SCr twice normal for patient age
 Few prospective pediatric studies
 Retrospective studies assess AKI causes
 Control group without AKI not assessed to determine risk factors
for AKI
Pediatric AKI Epidemiology until 2002:
What was Out There?
 Most original data all single center
 Predate current ICU technology and practice
 Predate recent disease therapies
 Bone marrow transplantation
 Cardiac transplantation
 Congenital heart surgery
 Cite Hemolytic-Uremic Syndrome/primary renal disease
as most common causes
 Most articles after 1995 are literature review
Patient Selection
 Reviewed all admissions to Texas Children’s Hospital from
January 1998 through June 2001
 Selected patients <20 years of age with ARF listed as
diagnosis on discharge or death summary
 Reviewed list and defined ARF as GFR by Schwartz < 75
ml/min/1.73m2 (n=254)
Most Common ARF Causes





ATN-Dehydration (21%)
Nephrotoxic drugs (16%)
Sepsis (11%)
Unknown (14%)
Primary Renal Disease (7%)
Patient Survival
 176/254 patients (70%)
 110/185 patients with ICU care (60%)
 43/77 patients receiving renal replacement therapy (56%)
Pediatric AKI Epidemiology
Author
Williams
Hui-Stickle
Akcan-Arikan
Year
Time
span
2002
19781998
2005
19992001
2007
20052006
Cohort
AKI Cause
All hospital
1978-88: HUS 38%,
Oncology 8%
1988-98: HUS 22%
Oncology 17%
All hospital
Ischemic 21%
Nephrotoxins 16%
Primary Renal 7%
PICU
Pneumonia (33%)
SIRS/sepsis(27%)
Cardiogenic (10%)
Pediatric AKI Risk Factors
 Few comparative data of populations with versus without AKI
to determine who is truly at risk
 Most data examine only patients with AKI and report causes
(previous slides)
AKI Risk Factors – Assessment Issues
 Retrospective aminoglycoside




study
AKI defined as 50% decrease
Heme/Onc and Pulm with
highest AKI
Surgery with lowest AKI
Heme/Onc and Pulm assessed
SCr significantly more often
than Surgery
50
40
30
%AKI
20
10
0
Peds
H-Onc
Pulm
Surg
1
0.9
0.8
0.7
0.6
0.5
#SCR per days treated
0.4
0.3
0.2
Zappitelli and Goldstein, submitted
Other
0.1
0
Peds
H-Onc
Pulm
Surg
Other
Pediatric AKI Risk Factors:
The Critically Ill Patient
 Highest risk for AKI development
 AKI now results from other systemic illness or its treatment
and not from primary kidney disease
 Most pediatric AKI studies focus on patients who receive
RRT
 More recent studies compare patients with AKI versus
without AKI
 Single-center, prospective observational study over one
year (2000-2001)
 Pediatric ICU population
 3 days to 18 years of age
 AKI defined as doubling SCr
 Doubling of upper limit of normal
 Doubling of PICU admission SCr
 True “baseline” pre-PICU SCr not assessed
 CKD patients: AKI defined as 25% increase in SCr
 1047 admissions
 Exclusions for patient age,
prematurity, decision to
withhold care, pregnancy
 4.5% AKI rate
 Risk factors
 Thrombocytopenia
 Older age
 Hypoxemia
 Hypotension
 Coagulopathy
 Increased PRISM and
PELOD scores also AKI
risk factors
Is All AKI Created Equal?
 Recent adult patient data demonstrate
 Small SCr rises associated with mortality
 AKI associated with mortality and length of hospitalization
 AKI is now recognized as risk factor for poor outcome,
independent of severity of illness
AKI Severity and Outcome
Chertow GM et al: J Am Soc Nephrol, 2005
All AKI is NOT Equal
 Multidimensional classification system is needed to
 Grade AKI severity
 Follow changes in kidney function
 Standardize AKI as a hard outcome measure
AKI RIFLE Criteria: ADQI II
 Prospective single center observational study
 PICU patients receiving mechanical ventilation and vasoactive
medications
 AKI defined by a pediatric modified RIFLE criteria (pRIFLE)
 pRIFLEmax defined as highest pRIFLE stratum achieved at 14 days of
PICU admission or patient discharge, whichever came first
 eCCl determined by Schwartz
formula
 Baseline eCCl from three
months before PICU
 100 ml/min/1.73m2 if no data
available
 pRIFLE differs from RIFLE in
 Oliguria duration
 RIFLE-F limit eCCl
AKI occurred early in PICU admission
• 82% of AKI patients attained their initial RIFLE stratum in the first 7 days.
Initial RIFLE R
N=76
Initial RIFLE I
N=31
3/76 (4%) RIFLEmax F
12/31 (39%) pRIFLEmax F
“Persistent” AKI on admission
Biomarkers
A biologic characteristic that is measured and evaluated objectively as an
indicator of normal biologic processes, pathogenic processes, or
pharmacologic response to therapeutic intervention. Hewitt et al, JASN, 2004
 imaging test (renal ultrasound for kidney size)
 gene expression profiles for specific health or disease states
 proteinuria
 lipid profile
 metabolomic profiles
Why do we need biomarkers of AKI?
 Because AKI is important.
 Independent RF for mortality and longer LOS in critically ill children.
Ackan-Arikan et al, KI, 2007; Plotz et al, Intens Care Med, 2008
Independent RF for LOS in children having cardiac surgery.
Bernier et al, ASN, 2008
 Independent RF for longer LOS in children treated with aminoglycosides.
Zappitelli et al, CJASN, 2008; Zappitelli et al, ASN, 2007
 May be a RF for long-term abnormal renal function problems.
Askenazi et al, KI, 2006
Why do we need biomarkers of AKI?
 No treatment.
 Diagnosis based on SCr rise: 1 to 3 days after injury – failed past clinical trials.
 Several issues with SCr as a marker of GFR.
Utilities of biomarkers in AKI
 Early diagnosis
 Define severity of injury, monitor AKI course
 Define AKI subtypes & etiology (pre-renal, septic, nephrotoxic)
 Monitor response to AKI interventions
 Risk stratify for poor outcomes (dialysis need, CKD, mortality)
 Identify location of renal tubular injury
Devarajan &Williams, Seminars in Nephrol, 2007
What is an ideal biomarker?
Qualities






Accurate, reliable
Relatively non-invasive/acceptable to
patients
Rapidly measurable, standardized assay
Sensitive/specific with reproducible
cutoff values
Requires case definition: AKIN,
pRIFLE
Nguyen & Devarajan, Ped Nephrol, 2008
Phases of biomarker discovery: bench to
bedside
Phases of biomarker development
Validation
Translational
Discovery
Devarajan &Williams, Seminars Nephrol, 2007; Coca & Parikh, CJASN, 2008
Phase
Terminology
Phase 1
Preclinical Discovery
• Discover biomarkers in tissues or body fluids
• Confirm and prioritize promising candidates
Phase 2
Assay Development
• Develop and optimize clinically useful assay
• Test on existing samples of established disease
Phase 3
Retrospective Study
• Test biomarker in completed clinical trial
• Test if biomarker detects the disease early
• Evaluate sensitivity, specificity, ROC
Phase 4
Prospective Screening
• Use biomarker to screen population
• Identify extent and characteristics of disease
• Identify false referral rate
Phase 5
Disease Control
Action Steps
• Determine impact of screening on reducing
disease burden
Biomarker discovery in AKI: bench to
bedside
 NGAL:
 Expressed in proximal and distal nephron
 Binds and transports iron-carrying molecules
 Role in injury and repair
 Rises very early (hours) after injury in animals, confirmed in children having CPB
 IL-18:
 Role in inflammation, activating macrophages and mediates ischemic renal injury
 IL-18 antiserum to animals protects against ischemic AKI
 Studied in several human models
 KIM-1:
 Epithelial transmembrane protein, ?cell-cell interaction.
 Appears to have strong relationship with severity of renal injury
Biomarker studies in different
populations
 Cardiac surgery
 Critically ill patients
 Sepsis
 Nephrotoxin-treated patients
 Renal transplant
 General hospital population
Cardiac surgery: Known timing of AKI
NGAL: Children led the way!
Mishra et al, Lancet, 2005
SCr rise
48-72 hrs
Adults
Wagener et al, Anesthesiology, 2006
Not quite as good
Cardiac surgery
Parikh et al,KI, 2006
Children
Critical Illness: unknown timing of AKI
Parikh et al, JASN, 2005
SCr rise
Critically ill adults: retrospective. Landmark study.
IL-18
Critical illness population
The day of SCr rise:
Can biomarkers tell us WHO has “true AKI” versus who has volume depletion?
Predict lack of SCr return to normal within 48 hrs when taken at time of SCr rise
KIM-1
0.50
0.25
0.00
Sensitivity
0.75
1.00
NGAL
0.00
0.25
Area under ROC curve = 0.7692
0.50
1 - Specificity
0.75
1.00
Texas Children’s AKI Biomarker Study
 Previous published pediatric AKI biomarker reports from
homogeneous patients populations, many with primary renal
disease
 Prospective study of 150 patients admitted to TCH PICU who
received mechanical ventilation and/or vasoactive medications
 Outcome Measures
 pRIFLEmax at 14 days of ICU admission
 Persistent AKI (AKI that did not resolve in 48 hours
 Predict AKI prior to pRIFLE
Texas Children’s AKI Biomarker Study
 150 patients (enrolled in pRIFLE study)
 10 patients excluded from biomarker study for anuria or no
indwelling Foley
 Urine obtained at 2 PM for up to four days after study
enrollment
 NGAL (Devarajan)
 IL-18 (Edelstein)
 KIM-1 (Bonventre)
 pRIFLE creatinine calculated from Day 1 to Day 14 of
ICU admission
 140 patients’ urine samples available
 Mean age 6.3 years (1 year to 21 years)
 Mean ICU day of admission = 3 + 1.5 days
 pRIFLE
 No AKI:
 R:
 I:
 F:
24.3%
33.7%
22.1%
17.9%
6
4
2
uNGAL (ng/mg creatinine)
0
●
-3
●
-2
●
-1
●
0
●
1
●
2
Days from Day of first pRIFLE (Day 0)
AKI higher than controls from Day -2 to Day 2, p<0.05
Increase in NGAL to predict AKI: AUC=0.78
Increase in NGAL to predict persAKI: AUC=0.80
Control
R
Mean uIL18
I
F
Peak uIL18
0
100
200
300
Mean and Peak uIL18 (pg/ml)
400
All Patients
Non-septic
400
300
200
0
First uIL-18 (pg/ml)
100
Survivors
Non-survivors
excludes outside values
P<0.05
Pediatric AKI and Biomarkers: Conclusions
 Pediatric AKI is seen as a complication of other systemic
illness
 Earlier recognition and treatment of AKI sequelae may
improve outcome
 Active investigation/validation of urinary biomarkers
may lead to therapies to prevent or mitigate the effects of
AKI