Challenges in Establishing a Neonatal AKI Definition ... Services...Challenges in Establishing a Neonatal AKI Definition & ... • Highly dynamic GFR in 1st few weeks of ... •Cystatin

Challenges in Establishing a Neonatal AKI Definition &

Long Term OutcomesCherry Mammen MD, FRCPC, MHSc

Pediatric Nephrologist, BC Children’s HospitalClinical Assistant Professor

University of British ColumbiaVancouver, BC, Canada

Neonatal AKI SymposiumCleveland, Ohio

October 7th, 2017

No Disclosures or Conflicts of Interest

Objectives

• To describe a recently proposed standardized & staged neonatal AKI definition (modified KDIGO definition)

• To review challenges in defining neonatal acute kidney injury (AKI)

• Serum creatinine• Urine output (oliguria)• Other AKI biomarkers

• To discuss the risk of chronic kidney disease (CKD) in survivors of neonatal AKI

AKI Definition Evolution

• >35 definitions of “acute renal failure” before 2000• Kellum JA. Curr Opin Crit Care 2000

• 2004: RIFLE criteria created • Bellomo R et al. Crit Care 2004

• 2007: Acute Kidney Injury Network (AKIN) criteria• Mehta RL et al. Crit Care 2007

• 2007: Pediatric (pRIFLE) criteria• Akcan-Arikan A et al. Kidney Int 2007

• 2012: KDIGO classification• Kidney Int Supplements 2012

• 2013: Neonatal modified KDIGO definition • Jetton J et al. The Lancet (Child/Adolescent)

2017

Neonatal AKI Definition (modified KDIGO criteria)

KDIGO definition independently associated with poor outcomes

Jetton J et al. The Lancet (Child-Adolescent) 2017 (in-press)

Challenges with serum creatinine

• Marker of function (not injury)• Rises once 25-50% of renal function is lost

• Rises late (24-48 hrs after initial injury)• Does not differentiate the nature of injury• Affected by fluid status (dilution from fluid overload)

• AKI incidence/staging changes when corrected for fluid overload

• Type of measurement (Enzymatic vs Jaffe)• Jaffe method

• Interference with albumin & bilirubin• May overestimate Cr (0.2 mg/dL in ELBW infants)

Basu RK et al. Pediatr Crit Care Med 2013Allegaert K et al. J of Maternal-Fetal & Neonatal Med 2012

Issues with creatinine (NICU)

• Interference with maternal creatinine• Highly dynamic GFR in 1st few weeks of life • Low muscle mass population• Frequency of Cr monitoring varies from NICU to

NICU• What threshold of Cr rise should we use?

• Historically, absolute SCr >1.5-2 mg/dL has been used • Are these thresholds too strict?

• Should we be using lower %∆SCr thresholds (pRIFLE, AKIN, KDIGO)?

Neonatal serum creatinine trends

Is this AKI?

Creatinine trends (premature infants)

Divided by levels by gestational age

Gallini F Pediatr Nephrol 2000Guignard JP Pediatrics 1999

Is this AKI?

?Tubular reabsorption of SCr

100 umol/L= 1.1 mg/dL120 umol/L = 1.35 mg/dL

Failure to drop?

Gupta C et al. Pediatr Nephrol 2016

Failure to drop?

Proposes cutoffs and rates of decline as diagnostic of AKI

Gupta C et al. Pediatr Nephrol 2016

Cystatin C vs Serum Creatinine

• Advantages of Cystatin C• Independent of age, sex, weight, height• Minimal amounts cross placenta• Filtered by glomerulus, completely reabsorbed, & not

secreted• Superior to serum Cr in estimating GFR in neonates• However, 5x the cost of SCr ($3-5/assay)

Abitbol CL et al. J Peds 2014

Urine output issues

• What is the definition of “oliguria” in a neonate?• Minimum urine output 1 cc/kg/hr to remain in solute balance• Most standardized definitions define “oliguria” <0.5 cc/kg/hr

• How do we measure U/O accurately?• Many NICU patients do not have Foley catheters• How do we deal with the “poop situation” & “combos” on

flow sheets?

• When do we start the urine output clock in NICU pts?• Many infants do not urinate right away

• Does adding U/O criteria change AKI incidence/staging?

24 hours good time to start the clock?

Clark DA. Pediatrics 1977

• Retrospective NICU study (312 infants)• 48% pre-term, 24% 5-min APGAR <7• Overall mortality 12.8%

• U/O measured by weighing diapers q3hours• AKI defined by pRIFLE (eGFR & U/O criteria)• Patients divided into U/O thresholds (excluding 1st

24 hrs of life)• Group 1: >1.5 cc/kg/hr x 24 hrs • Group 2: 1-1.5 cc/kg/hr x 24 hrs• Group 3: 0.7-1 cc/kg/hr x 24 hrs• Group 4: <0.7 cc/kg/hr x 24 hrs

Bezerra CTM et al. Nephrol Dial Transplant 2013

pRIFLE criteria: AUC for mortality 0.689, addition of proposed U/O criteria 0.885Bezerra CTM et al. Nephrol Dial Transplant 2013

Adding U/O Criteria Changes IncidenceAWAKEN Study

Jetton J et al. The Lancet (Child-Adolescent) 2017 (in-press)

Incidence with creatinine criteria: 380/2022 (19%)Incidence with Cr or U/O criteria: 605/2022 (30%)

Neonatal AKI Biomarkers

Published norms for premature infants without AKI

Saeidi B et al. Pediatr Nephrol 2015

AKI biomarkers in premature infants during first week of life

Askenazi D et al. CJASN 2016

Table 5. Performance for biomarker value to acute kidney injury (maximum biomarker level for all, except those shown with ** which areminimum values)

(pg/mL) Ideal crude cutoff *Crude AUC*GA Adjusted

AUCAlbumin 3.69×107 0.62 0.68

CystatinC 1.26×106 0.65 0.68EGF** 5.9×102 0.68 0.70NGAL 4.5×105 0.63 0.67OPN 7.12×105 0.65 0.67

UMOD** 6.7×105 0.71 0.73Clusterin 3.1×105 0.62 0.67α-GST 4.33×102 0.68 0.68

* log transformed before analysis to normalize the distribution.

Take Home Points• Complex neonatal physiology makes creatinine a

flawed clinical biomarker of AKI• Oliguria is important to document with mortality

association• Ideal threshold (<1 vs 1.5 cc/kg/hr) not yet answered

• Modified KDIGO classification (both SCr & U/O) should be used consistently until improved definitions come out

• Urine AKI biomarkers in neonates are promising, but further work is needed

Long Term Neonatal Renal Outcomes

AKI CKD

What is Chronic Kidney Disease?

Kidney International 2013

eGFR = 0.413 x (height/Scr) if height is expressed in centimeters (Schwartz)

How Bad Is to Have Chronic Kidney Disease?

ANSWER: VERY BAD!!!!

Age-Standardized Rates of Death from Any Cause, Cardiovascular Events, and Hospitalization, According to the

Estimated GFR among 1,120,295 Ambulatory Adults.

Death CV Event Hospitalization

Go AS et al. N Engl J Med 2004

AKI & CKD Link in Adults is Strong

Greenberg JH et al. BMC Nephrology 2014

Limited neonatal AKI to CKD literature

• Anand SK J Pediatr 1978 • Mocan H Pediatr Nephrol 1991 • Shaw NJ Int J Cardiol 1991• Marks SD J Pediatr 2005• Polito C Clin Pediatr 1998• Abitbol CL Pediatr Nephrol 2003• Zwiers AJM Clin J Am Soc 2014• Bruel A Pediatr Nephrol 2016• Harer MW Pediatr Nephrol 2017 • Maqsood S Pediatr Nephrol 2017• Chaturvedi S Pediatr Nephrol 2017 (Review Article)

• Long term outcome of 20 ELBW infants with AKI • Peak SCr >2mg/dl > 48hrs and/or oliguria (<0.5 ml/kg/hr)

after 3rd DOL

• Mean age at F/U: 7.5 +/- 4.6 yrs (range 3.2-18.5 yrs)• Chronic renal impairment: eGFR <75 ml/min/1.73m2

• Also assessed for proteinuria (Urine PCR >0.2 mg/g), HTN (casual office BP), and renal size (U/S)

• 9/20 (45%) patients identified with low eGFR

• Rotterdam, Netherlands ECMO program• 423 neonates undergoing ECMO from

1992-2002• 65% Incidence of AKI (pRIFLE)• Median age of follow-up: 8.2 years

Zwiers AJM et al. Clin J Am Soc 2014

HTNn=25

proteinurian=17

eGFR<90 ml/min/1.73

m2n=5

No CKD or HTNn=115

ECMO~620

29% of neonates exposed to ECMO developed a sign of

chronic renal injury

Children with RIFLE scores injury and failure

4.3 times higher odds of CKD signs or HTN

• 34 VLBW infants followed up at 5 years of age • 20 with neonatal AKI and 14 without Neonatal AKI

• 9/34 (26%) had Cystatin C eGFR<90 mL/min/1.73 m2 (p = 0.25)• 7/20 (35%) with AKI • 2/14 (14%) without AKI

• At least one sign of CKD (p< 0.05)• 13/20 (65%) with AKI• 2/14 (14%) w/o AKI

VLBW neonates with AKI had a

4.5 times greater risk of CKD

Harer MW et al. Pediatr Nephrol 2017

Many other factors associated with CKD in neonates

Chaturvedi S et al. Pediatr Nephrol 2017

Albuminuria:

OR = 1.81 (1.19, 2.77)

ESRD:

OR = 1.58 (1.33, 1.88)

Low eGFR:

OR = 1.79 (1.31, 2.45)

White SL et al. Am J Kidney Dis 2009

How should we follow-up?

Carmody JB et al. Pediatrics 2013

Take Home Points• Survivors of neonatal AKI may be at high risk of CKD• Even without AKI, babies with LBW/IUGR and/or

prematurity are at risk of CKD• Long-term renal surveillance is needed

• Nephrologists & neonatologists need to work together to make this happen

• What can “we” do for now?• Improve knowledge translation (spread the word)• Better recognition and recording of AKI episodes• Concentrate on AKI awareness & prevention• Design better larger scale prospective studies

Questions/CommentsDr. Cherry Mammen

Clinical Assistant Professor (UBC)Pediatric Nephrologist/Director of Dialysis

BC Children’s HospitalVancouver, BC, Canada

Email: [email protected]

• 244 Neonatologists (73% US based) & 131 Nephrologists (75% US based)

• Based on 3 NICU cases scenarios representing KDIGO AKI Stages 1, 2, & 3

• 98% of nephrologists vs 51% of neonatologists were aware of published neonatal AKI definitions

• Stage 1: 62% of neonatologists diagnosed AKI • Stage 2: 76% of neonatologists diagnosed AKI• Stage 3: 99% of neonatologists diagnosed AKI

Kent AL et al. Am J Perinatol 2017 (Epub ahead of print)

Bezerra CTM et al. Nephrol Dial Transplant 2013

Cystatin C vs Serum Creatinine

•SCr based equations underestimated inulinGFR by >20%

•Seen at all gestational ages

•Cystatin C is superior to SCr in neonates in estimating GFR

Abitbol CL et al. J Peds 2014