부산대학교 도서관

Patients with a Fontan circulation are at risk of renal dysfunction. We analyzed cross‐sectional data in pediatric and adult Fontan patients in order to assess the accuracy of commonly used serum creatinine‐based methods in estimating glomerular filtration rate (GFR). A total of 124 Fontan patients (58 children, 66 adults) were enrolled across three study centers. Measurement of GFR (mGFR) using in vivo 99mTc‐DTPA clearance was performed. Various serum creatinine‐based equations were used to calculate estimated GFR (eGFR). Mean mGFR was 108 ± 28 mL/min/1.73 m2in children and 92 ± 20 mL/min/1.73 m2in adults. Fourteen children (25%) and 28 adults (45%) had an mGFR <90 mL/min/1.73 m2. There was no significant correlation between mGFR and eGFR (Schwartz) in children (r= 0.22, P= .1), which substantially overestimated mGFR (bias 50.8, 95%CI: 41.1‐60.5 mL/min/1.73 m2, P< .0001). The Bedside Schwartz equation also performed poorly in the children (r= 0.08, P= .5; bias 5.9, 95%CI: −2.9‐14.6 mL/min/1.73 m2, P< .0001). There was a strong correlation between mGFR and both eGFR (CKD‐EPI) and eGFR (MDRD) in adults (r= 0.67, P< .0001 in both cases), however, both methods overestimated mGFR (eGFR(CKD‐EPI):bias 23.8, 95%CI: 20‐27.6 mL/min/1.73 m2, P< .0001; eGFR (MDRD):bias 16.1, 95%CI: 11.8‐20.4 mL/min/1.73 m2, P< .0001). None of the children with an mGFR <90 mL/min/1.73 m2had an eGFR (Schwartz) <90 mL/min/1.73 m2. Sensitivity and specificity of eGFR (CKD‐EPI) and eGFR (MDRD) for mGFR <90 mL/min/1.73 m2in adults were 25% and 92% and 39% and 100%, respectively. This study identifies the unreliability of using creatinine‐based equations to estimate GFR in children with a Fontan circulation. The accuracy of formulas incorporating cystatin C should be further investigated and may aid noninvasive surveillance of renal function in this population.