Middle-molecules (MM) are not monitored in children on hemodialysis (HD), but are accumulated and increase the risk of cardiovascular disease and mortality. Molecular properties of Cystatin C (CyC), 13 kDa, potentially make it a preferred MM marker over Beta-2-Microglobulin (B2M), 12 kDa. We compared CyC and B2M kinetics to investigate if CyC can be used as preferred MM marker. CyC (mg/L) and B2M (μg/mL) were measured in 21 low-flux HD sessions in seven children. Blood samples were taken at HD start (pre), 1 and 2 hours into HD and at end of HD (post) for all sessions and 60 minutes after the first HD (Eq). PreCyC (9.85 ± 2.15) did not differ (P > 0.05) from postCyC (10.04 ± 2.83). PostB2M (38.87 ± 7.12) was higher (P < 0.05) than preHD B2M (33.27 ± 7.41). There was no change in CyC at 1 and 2 hours into HD, while B2M progressively increased. CyC or B2M changes did not significantly correlate with spKt/V (2.09 ± 0.86), ultrafiltration (4.61 ± 1.98%) or HD duration (218 ± 20 minutes). EqCyC was not different from postCyC (11.07 ± 3.14 vs. 10.71 ± 2.85, P > 0.05), while EqB2M was lower than postB2M (36.48 ± 7.68 vs. 41.09 ± 8.99, P < 0.05). MMs as represented by B2M and CyC are elevated in children on standard HD. Intensified HD modalities would be needed for their removal. B2M is affected by the dialytic process with a rise during HD independent of ultrafiltration and decrease 1 hour after, while CyC remains unchanged. We suggest that CyC be used as preferred marker of MM removal and as a marker of adequacy of intensified HD regimens.

A 9-month old girl presented to the emergency room with brown urine and generalized edema. She had been born at full term with normal growth and development. Her parents were unrelated and had no other children. She had been in good health until she developed brown urine 4 days prior to presentation and was noted to have edema 1 day prior to being seen in emergency department. Her physical exam was remarkable for moderate generalized edema, mild pallor, and irritability. Vital signs were within normal limits except for elevated blood pressure (115/63 mmHg). Her urine specimen was turbid and brown. Microscopic exam showed red blood cells (RBC) too numerous to count, 2–5 white blood cells (WBC), and 5–10 squamous epithelial cells per high power field. No RBC casts were seen. Urine protein/creatinine ratio was 17. Renal function panel revealed normal electrolytes, elevated serum creatinine at 0.49 mg/dl (estimated glomerular filtration rate Schwartz 63 ml/min/1.73 m), and low serum albumin at 1.9 g/dl. Complete blood count showed mild anemia with hemoglobin 8.9 g/dl, platelet count 273,000/mm, and WBC 13,900/ mm. Renal ultrasound showed bilaterally enlarged kidneys with increased cortical echogenicity. Chest X-ray was within normal limits. The infant was admitted to the hospital.

Online clearance (OLC) monitor measures conductivity difference between dialysate entering and leaving the dialyser. Derived ionic dialysance (ID) represents effective urea clearance from which Kt/V is calculated, allowing Kt/V monitoring at every treatment without blood sampling. We tested ID accuracy in children and provide recommendations for its use. Using Fresenius machines 2008 K with built-in OLC monitors, we studied 45 hemodialysis (HD) sessions and 168 calculated Kt/V results in 11 patients. Urea distribution volume (V), needed to calculate Kt/V from ID, was estimated using three methods: Mellits and Cheek (MC), KDOQI recommended total body water nomograms (TBWN) and OLC-derived independent from tested HD sessions. Reference spKt/V from pre-and post-HD BUN (Daugirdas) was compared with Kt/V calculated from ID using three different estimated V's. ID was accurate in calculating Kt/V in children when V derived from OLC was used (P = 0.42), with absolute error 0.14 +/- 0.12. If TBWN-derived V was used, Kt/V was consistently underestimated by 0.32 +/- 0.22. TBWN-derived V can still be recommended for use with OLC for monitoring trend in Kt/V, if underestimation of spKt/V of average 0.3 is accounted for. MC-derived V results in even greater underestimation of spKt/V and therefore cannot be recommended for use with OLC.