AKI in Pediatrics

1.      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

2. OBJECTIVES • The Problematic Definition of ARF • 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

3. 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”

4. 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

5. 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……?

6. 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

7. 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

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

9. EPIDEMIOLOGY

10. 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)

11. 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.

12. 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

13. Pediatric AKI Literature: EpidemiologyWhat’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

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

15. 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

16. 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

17. 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

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

19. 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

20. 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

21. BIOMARKERS

22. 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) • Neutrophilgelatinase-associated lipocalcin (NGAL) • IL-18 • Cystatin C

23. Nguyen- pedneph 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

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

25. 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

26. 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

27. 225 200 175 150 125 100 75 50 25 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 Detection of Urinary NGAL by ELISA 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

28. 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)

29. 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

30. RRT OUTCOME

31. 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

32. 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

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

34. 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

35. 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

36. 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

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

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

39. 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

40. 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

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

42. 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

43. 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?

44. 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?

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