AKI in Pediatrics Patrick D. Brophy MD Associate Professor University of Iowa- Carver College of Medicine Dept. of Pediatrics Division of Nephrology, Hypertension, Dialysis & Transplantation

OBJECTIVES The Problematic Definition of ARF The Transition New concepts/old habits The Transition Acute Renal Failure Acute Kidney Injury Epidemiology- developing and developed countries HUS to ATN Biomarkers? How are we doing NGAL, IL-18 and creatinine RRT Treatment Outcomes Update on variables in the PICU Modalities of choice? PD, HD or CRRT--- What modality, when and why

The Problematic Definition of ARF The Conceptual Definition of Acute Renal Failure: “Sudden loss of renal function resulting in the loss of the kidneys’ ability to regulate electrolyte and fluid homeostasis”

The Problematic Definition of ARF Pediatric AKI definition: a moving target Infants Cr in the first few weeks of life may reflect maternal values Children Low baseline Cr makes 0.2-0.3 changes in Cr significant Varying muscle mass Adolescents Similar to adults

The Problematic Definition of ARF Over 30 published ARF definitions All based on increased serum creatinine levels Despite extensive adult hospitalized patient study over the past 50 years Widely varying spectrum dependent upon study aims and hypothesis Severe (ARF requiring dialysis) Modest (serum creatinine increase of 0.3 mg/dl) Why is this important……?

The Problematic Definition of ARF The lack of a uniform ARF definition has prevented optimal ARF outcome research One study’s ARF is another study’s lab error (or maybe not) Inherent problems with SCr as ARF marker Does not differentiate the nature and type of renal insult site of renal insult Changes in SCr may lag changes in GFR and may be a very late indicator of renal injury Dialysis removal negates marker effectiveness

The Ideal Disease Definition Would ascertain disease presence guide nature and timing of diagnostic and therapeutic interventions help determine prognosis Should incorporate clinical signs and symptoms alterations in reproducible biological markers

Pediatric Modified RIFLE--definition GFR per Schwartz equation: GFR= Ht (cm) X constant / serum creat (mg/dl) Pediatric Modified RIFLE Criteria CrCl Urine output Risk GFR decrease by 25% <0.5ml/kg/hour for 8 hours Injury GFR decrease by 50% 16 hours Failure GFR decrease by 75% or GFR<35ml/min/1.73m 2 <0.3 ml/kg/hour for 24 hours or anuric fo r 12 hours Loss Persistent ARF > 4 weeks End stage End Stage Renal Disease (>3 months) Ackan-Arikan et al: Kid Int 2007

EPIDEMIOLOGY

Epidemiology-The Pediatric Patient with AKI IS NOT a small adult 0 days to 21+ years 2 kg to 200 kg Primary conditions Congenital heart disease Inborn errors of metabolism Sepsis with multi-organ involvement Bone marrow and solid organ transplantation Children develop and die of MODS early in ICU course Maximum number of organ failures occurs within 72 hours of ICU admission (87% of patients) 88.4% of deaths occur within 7 days of MOSF diagnosis (Proulx et al: Crit Care Med 22:1025, 1994)

Epidemiology-Incidence AKI A retrospective analysis reported an incidence rate of AKI of 2.7% (defined as need for dialysis) in children undergoing cardiopulmonary bypass surgery [Picca NDT 1995]. The prospective study validating the Pediatric Logistic Organ Dysfunction (PELOD) score in pediatric intensive care units (PICU), the incidence of AKI (defined as serum creatinine levels above 55 mol/L to 140 mol/L depending on age of the child) was 129 per 1000 admissions [Leteurtre lancet 2003]. PICU prospective trial reported an incidence rate of AKI of 44.7/1000 admissions [Bailey reanimation 2005]. In the face of the lack of common defining terms of pediatric AKI, clear incidence and prevalence data is difficult to establish based on the literature.

Pediatric AKI - Incidence of cases requiring RRT 227 cases in the years 84-91. Yorkshire UK Incidence: 0.8/yr/100.000 total population neonate-infant: 19.7/yr/100.000 age related population 1-4 years: 5.9/yr/100.000 age related population 5-15 years: 1.5/yr/100.000 age related population children: 3.9/yr/100.000 age related population 1/5 of the adult incidence

Pediatric AKI Literature: Epidemiology What’s 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 and other primary renal disease as most common causes Most articles after 1995 are literature review

Epidemiology/Etiology 11% - 25(20.3%) Sepsis 37(30%) Infections (malaria) 19(7.7%) Genetics diseases 11(4.4%) Hepatic/intestinal transplantation 33(13.3%) 17(13.4%) renal Burkitt lymphoma Hematology /oncology 43(17.3%) 11(13.4%) Post cardiac surgery Renal ischemia or nephrotoxic drugs in the context of: 20(8.0%) 7(8.7%) 7(5.7%) Urology 19(23.8%) Acute tubular necrosis 9(3.6%) 18(22.5%) 10(8, 1%) AGN 3(1.2%) 25(31%) 2(1.6%) HUS Causes 254 80 123 Number of patients (%) United States India Kenya Location Hui-Stickle et al[ajkd 2005] Arora et al [ped neph 1997] Oluwu et al[KI 2004] Study

Pediatric AKI: Recent Epidemiology 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 AKI as GFR by Schwartz < 75 ml/min/1.73m2 (n=254) Stickle SH et al: Am J Kid Dis 45:96-101, 2005

Pediatric AKI: Recent Epidemiology Data Reviewed Retrospective chart review for the following data: Patient age (years) and size (kg) Disease/condition leading to ARF Pediatric Renal Service consult obtained (yes/no) Corrected GFR (ml/min/1.73m2) by Schwartz formula nadir during ARF course GFR at time of Pediatric Renal Service consult Renal replacement therapy required (yes/no) ICU care required (yes/no) pressors required (yes/no) ICU length of stay (days) Survival defined as discharge from hospital Stickle SH et al: Am J Kid Dis 45:96-101, 2005

Pediatric AKI: Recent Epidemiology Most Common ARF Causes ATN-Dehydration (21%) Nephrotoxic drugs (16%) Sepsis (11%) Unknown (14%) Patient Survival 176/254 patients (70%) 110/185 patients with ICU care (60%) 43/77 patients receiving renal replacement therapy (56%) Stickle SH et al: Am J Kid Dis 45:96-101, 2005

Pediatric AKI: Recent Epidemiology Stickle SH et al: Am J Kid Dis 45:96-101, 2005

Pediatric AKI: Recent Epidemiology Average Length of ICU stay, days RRT – Renal Replacement Therapy Stickle SH et al: Am J Kid Dis 45:96-101, 2005

Pediatric AKI: Recent Epidemiology Renal Function at Hospital Discharge 116/176 (66%) survivors completely recovered 50/176 (29%) had improved renal function or chronic renal insufficiency 11/176 (5%) required renal replacement therapy Stickle SH et al: Am J Kid Dis 45:96-101, 2005

BIOMARKERS

Biomarkers for Acute Kidney Injury Ideally AKI would have a biomarkers like myocardial infarction (i.e. troponin-1) Currently no Troponin-I like marker to identify the site or severity of injury, although various markers are being evaluated Kidney Injury Molecule (KIM-1) Neutrophil gelatinase-associated lipocalcin (NGAL) IL-18 Cystatin C

Nguyen- ped neph 2007 Three hypothetical receiver-operating characteristic (ROC) curves are shown. The blue (straight) line represents a biomarker with an area under the curve (AUC) of 0.5, which indicates a result that is no better than expected by random chance. The red (middle) curve yields an AUC of about 0.75, which is generally considered a good biomarker. The green (top) curve gives an AUC of approximately 0.9, which would represent an excellent biomarker

Nguyen- ped neph 2007  Current status of promising acute kidney injury (AKI) biomarkers in various clinical situations None Not Tested Not tested Intermediate Urine KIM-1 Absent IL-18 Abbotta Early NGAL Dade-Behring Plasma Cystatin C Biositea Commercial Test? Kidney Transplant Sepsis or ICU Contrast Nephropathy Cardiac Surgery Sample Source Biomarker Name NGAL neutrophil gelatinase-associated lipocalin, IL-18 interleukin 18, KIM-1 kidney injury molecule 1 aIn development

Methods Infants had urinary NGAL assessed at frequent intervals after undergoing cardiac bypass for congenital heart surgery The primary outcome variable was the development of ARF, defined as a 50% or greater increase in serum creatinine Other data collected included age, gender, bypass time, previous surgery, urine output, urine creatinine, urine NGAL, length of ICU stay, complications, and death Mishra J et al: Lancet 2005

Incidence and Timing of AKI Number of Patients Patients with ARF No ARF (n=51) ARF (n=20) 24 48 72 Post CPB Time (hours) Using serum creatinine, the diagnosis of ARF can be made only after 24-72 hours post CPB Mishra J et al: Lancet 2005

Detection of Urinary NGAL by ELISA 15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 25 50 75 100 125 150 175 200 225 Serum Creatinine Rise Urine NGAL (ng/ml) ARF (n=20) No ARF (n=51) 2 4 6 8 12 24 36 48 60 72 84 96 108 120 Post CPB Time (hours) Urine NGAL is upregulated 15-fold within 2 hours after CPB in patients who later develop ARF Mishra J et al: Lancet 2005

Biomarkers in Pediatric AKI 140 children evaluated for AKI (heterogeneous population) In those with AKI 6 x greater of uNGAL on day 0 of pRIFLE as compared to no AKI IL 18 elevation was noted on day 0 of pRIFLE as compared to no AKI Washburn KK, et al Nephrol Dial Transpln 2007 (epub) Zappitelli M et al Crit Care 2007 (epub)

Summary A consistent AKI classification system may allow for more reliable assessment of the effect of AKI on patient outcome pRIFLE Waiting for classic signs of kidney failure may lead to unnecessary morbidity and mortality, early biomarkers are critical to assess for AKI risk Combinations of biomarkers early NGAL, prognostic and differential intermediate markers Help with guidance towards early initiation and which type of therapy

RRT OUTCOME

Renal Replacement Therapy in the PICU: Pediatric Outcome Literature Few pediatric studies (all single center) use severity of illness measure to evaluate outcomes in pediatric RRT: Lane noted that mortality was greater after bone marrow transplant who had > 10% fluid overload at the time of HD initiation Smoyer2 found higher mortality in patients on pressors Faragson3 found PRISM to be a poor outcome predictor in patients treated with HD Zobel4 demonstrated that children who received CRRT with worse illness severity by PRISM score had increased mortality Did not stratify by modality 1. Bone Marrow Transplant 13:613-7, 1994 2. JASN 6:1401-9, 1995 3. Pediatr Nephrol 7:703-7, 1994 4. Child Nephrol Urol 10:14-7, 1990

CRRT and Outcome in Children Retrospective review of all patients who received CVVH(D) in the Texas Children’s Hospital PICU from February 1996 through September 1998 (32 months) Pre-CVVH initiation data: Age Primary disease leading to need for CVVH Co-morbid diseases Reason for CVVH Fluid intake (Fluid In) from PICU admission to CVVH initiation Fluid output (Fluid Out) from PICU admission to CVVH initiation GFR (Schwartz formula) at CVVH initiation Goldstein SL et al: Pediatrics 2001 Jun;107(6):1309-12

Percent Fluid Overload Calculation [ ] Fluid In - Fluid Out ICU Admit Weight % FO at CVVH initiation = * 100% Goldstein SL et al: Pediatrics 2001 Jun;107(6):1309-12

CRRT and Outcome in Children PRISM scores at PICU admission and CVVH initiation calculated by same nurse PICU Course Data: Maximum number of pressors used Pressors completely weaned (y/n) Mean Airway Pressure (Paw) at CVVH initiation and termination ICU length of stay (days) CVVH complications Outcome (death or survival) Goldstein SL et al: Pediatrics 2001 107:1309-12

CRRT and Outcome in Children 22 pt (12 male/10 female) received 23 courses (3028 hrs) of CVVH (n=10) or CVVHD (n=12) over study period. Overall survival was 41% (9/22). Survival in septic patients was 45% (5/11). PRISM scores at ICU admission and CVVH initiation were 13.5 +/- 5.7 and 15.7 +/- 9.0, respectively (p=NS). Conditions leading to CVVH (D) Sepsis (11) Cardiogenic shock (4) Hypovolemic ATN (2) End Stage Heart Disease (2) Hepatic necrosis, viral pneumonia, bowel obstruction and End-Stage Lung Disease (1 each) Goldstein SL et al: Pediatrics 2001 107:1309-12

CRRT and Outcome in Children Lesser % FO at CVVH (D) initiation was associated with improved outcome (p=0.03) Lesser % FO at CVVH (D) initiation was also associated with improved outcome when sample was adjusted for severity of illness (p=0.03; multiple regression analysis) Goldstein SL et al: Pediatrics 2001 107:1309-12

Fluid Overload as a Risk Factor N=113 *p=0.02; **p=0.01 Foland et al, CCM 2004; 32:1771-1776

Fluid Overload as a Risk Factor Kaplan-Meier survival estimates, by percentage fluid overload category Gillespie et al, Pediatr Nephrol (2004) 19:1394-1999

The Prospective Pediatric CRRT (ppCRRT) Registry No single pediatric center cares for enough CRRT patients annually to analyze the effect of more than a few variables on patient outcome

ppCRRT Experience First patient enrolled on 1/1/01 ~400 patients entered into database Currently 13 active pediatric centers Texas Children’s Boston Children’s Seattle Children’s UAB University of Michigan Mercy Children’s, KC Egleston Children’s, Atlanta All Children’s, St. Petersburg DC Children’s Columbus Children’s Packard Children’s, Palo Alto DeVos Children’s, Grand Rapids Cleveland Clinic

ppCRRT MODS Data: Clinical Variables Goldstein SL et al: Kidney International 2005

ppCRRT MODS Data: Other Analyses 77% of non-survivors die within 3 weeks of ICU admission Survival rates similar by CRRT modality (H 57%), (DF 53%), (HD 50%) Survival rates similar for patients on: 0-1 (53%), 2 (54%) or 3+ (39%) pressors Survival rates better for patients with: <20% FO (59%) versus >20% FO (35%) at CRRT initiation (p<0.001) Goldstein SL et al: Kidney International 2005

OUTCOMES Impacting on AKI demands that we overcome barriers perceived or real in order to optimize therapies Technology can be adapted, we wouldn’t hold ventilation if the patient’s clinical and biochemical measures deemed it necessary---so why does this happen in the case of AKI? Access? Fear? Failure to appreciate the consequences of with-holding therapy? Knowing when to start? (the great debate) Knowing what modality to use (availability?) Available data?

FUTURE WHEN HAS SYMPTOMATIC THERAPY FAILED: TIMING OF RRT INITIATION IN PEDIATRIC AKI CAN WE INTERVENE BEFORE USING BIOMARKERS AS A GUIDE? WHAT THERAPY TO USE? WHAT VARIABLES TO ASSESS ALONG THE WAY?

Thanks Stu Goldstein MD -slides Tim Bunchman MD ppCRRT members PICU and Nephrology nursing staff