1. Pediatric CRRT: The Prescription Stuart L. Goldstein, MD Associate Professor of Pediatrics Baylor College of Medicine

2. What’s in a CRRT Prescription? Indication (Why? Who? When?) Technical Aspects (What?)  Fluids (Symons)  Anticoagulation (Brophy)  Access (Hackbarth) CRRT Delivery (How?)  Blood pump flow rates  Modality  Priming  Dose

3. Why CRRT in AKI? Critically ill patient Advantages  Slower blood flows  Slower UF rates  UF rates can be prescriptive (versus PD)  Adjust UF rates with hourly patient intake  Increased cytokine (bad humors) removal? Disadvantages  Increased cytokine (good humors) removal?  Non-dialysis personnel with many other bedside responsibilities required to monitor circuit

5. When Should CRRT Be Started? Standard AKI criteria not responsive to medical therapy OR only preventable with limiting adequate nutrition  Uremia  Hyperkalemia  Acidosis  Fluid Overload Prevention of worsening fluid overload?

6. Timing of Pediatric RRT No adequate definition for “timing of initiation” Absence of a generally accepted, validated and applied AKI definition has impeded the adequate investigation of this question The decision to initiate RRT affected by  Strongly held physician beliefs  Patient characteristics  Organizational characteristics

7. Retrospective evaluation of 226 children who received RRT for AKI from 1992-1998 Pressor use surrogate marker for patient severity of illness Survival defined at PICU discharge

9. 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

10. Percent Fluid Overload Calculation % FO at CVVH initiation = [ Fluid In - Fluid Out ICU Admit Weight ] * 100% Fluid In = Total Input from ICU admit to CRRT initiation Fluid Out = Total Output from ICU admit to CRRT initiation

11. 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)

12. Survival curve demonstrates that nearly 75% of deaths occurred less than 25 days into the ICU course

13. 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)

14. N=113 *p=0.02; **p=0.01

15. Group Fluid Overload Hazard Ratio (95% CI)P All Subjects <10%1 0.002 >10%3.02 (1.50-6.10) N = 77

16. Kaplan-Meier survival estimates, by percentage fluid overload category

17. The Evolution of Idea to Practice Paradigm Single center study Registry Randomized Trial

18. Prospective Pediatric CRRT (ppCRRT ) Registry: Phase 1 Design Collect prospective data from 10 pediatric centers treating 15 to 20 patients annually (376 patients over 5 years) Each center follows own institutional practice  Patient selection  Initiation and termination  Anti-coagulation protocols  Convection versus diffusion versus hemodiafiltration  Fluid composition

19. Seven center study from the ppCRRT Registry 116 patients with MODS PRISM 2 score used to assess patient severity of illness Survival defined at PICU discharge

20.  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 (40%) at CRRT initiation (p<0.001)

21. The PCRRT Prescription: How? Blood pump flow rates Membranes Priming Modality Dose UF rates

22. Blood flow rates vary by patient size  Mean 5 ml/min/kg  CRRT clearance not limited by Qb 50% of ppCRRT patients received some convection

24. Pediatric CRRT Circuit Priming Heparinized (5000 units/L) for most patients Smaller patients require blood priming to prevent hypotension/hemodilution  Circuit volume > 10-15% patient blood volume  Packed RBCs Citrated – low ionized calcium Acid load Potassium load

25. Bradykinin Release Syndrome Mucosal congestion, bronchospasm, hypotension at start of CRRT Resolves with discontinuation of CRRT Thought to be related to bradykinin release when patient’s blood contacts hemofilter  Most common with AN-69 membranes Exquisitely pH sensitive

26. Technique Modifications to Prevent Bradykinin Release Syndrome Buffered system  THAM, CaCl, NaBicarb to PRBCs Bypass system  prime circuit with saline, run PRBCs into patient on venous return line Recirculation system  recirculate blood prime against dialysate

27. PRBC Waste

28. D Recirculation Plan: Qb 200ml/min Qd ~40ml/min Time 7.5 min Normalize pH Normalize K +

29. Does Modality Make A Difference? Equal clearance of smaller molecules Middle and large molecule clearance enhanced by convection

30. Creatinine 113 D Urea 60 D Glucose 180 D Vit. B 12 1,355 D  2-M 11,800 D Albumin 66,000 D IgG 150,000 D Membrane Selectivity Courtesy of J. Symons

31. Clearance: Convection vs. Diffusion

32. Solute Molecular Weight and Clearance Solute (MW)Sieving Coefficient Diffusion Coefficient Urea (60)1.01 ± 0.051.01 ± 0.07 Creatinine (113)1.00 ± 0.09 1.01 ± 0.06 Uric Acid (168)1.01 ± 0.04 0.97 ± 0.04* Vancomycin (1448)0.84 ± 0.10 0.74 ± 0.04** *P<0.05 vs sieving coefficient **P<0.01 vs sieving coefficient

33. Flores FX et al: CRRT 2006 abstract

36. ppCRRT Pediatric Sepsis Outcome Data 57/102 (56%) pts survived. Ventilated pts had similar survival rate as non- ventilated pts (53% vs. 68%, p=0.1). There was no significant difference in the survival rate among CRRT modalities. Tendency toward better survival with convective therapies Flores FX et al: CRRT 2006 abstract

37. Survival Based on CRRT Modality? Confounded  Center  Timing of initiation  Sepsis definition not standardized Suggestive  If all else equal, why not convect? Flores FX et al: CRRT 2006 abstract

38. Dialysate/ Ultrafiltration Rates The UF rate/plasma flow rate [=BFRx(1-HCT)] ratio should < 0.35-0.4 in order to avoid filter clotting (Golper AJKD 6: 373-386,1985) Dialysate flow rates ranging from 20-30 ml/min/m 2 (~2000ml/1.72m 2 /hr) are usually adequate (experiential but consistent with adult data)

40. Median survival  Group 1 (19 days)  Group 2 (33 days)  Group 3 (46 days) Groups 2 and 3 with longer survival than Group 1

41. Minimum UF rates > 35 ml/kg/hr Translates to approximately 2000ml/1.73m2/hour for children

42. Dose: Pediatric CRRT No published data to suggest an adequate or optimal CRRT dose in children Small molecule clearance and electrolyte homeostasis is generally easy to achieve Is more better?  Nutrition balance (what are we removing that we’d like to leave behind?)