Variability in Cystatin C- and Creatinine-estimated Glomerular Filtration Rate in Adults With Spina Bifida

To determine if serum cystatin C, which does not depend on muscle mass, is a potential alternative measure of kidney function in individuals with Spina Bifida and lower urinary tract dysfunction. Assessment using serum creatinine to estimate glomerular filtration rate is often compromised in this population because of decreased muscle mass from paralysis, which varies over spinal lesion levels and sub-phenotypes. We hypothesized creatinine-based formulas overestimate glomerular filtration rate at higher lesion levels relative to cystatin C-based formulas, and cystatin C-based equations provide stable estimates.

We retrospectively analyzed data from 155 adults with Spina Bifida. Estimated glomerular filtration rates from 5 equations (Cr-eGFR-2009, Cr-eGFR-2021, CysC-eGFR-2012, Zappitelli-CysC-eGFR, and CrCysC-eGFR-2021) were compared across 5 lesion levels and 2 Spina Bifida sub-phenotypes (myelomeningocele vs non-myelomeningocele).

Creatinine-based estimated glomerular filtration rate increased with higher lesion levels (both P <.001), overestimating kidney function by a median of 12 mL/min/1.73 m² in thoracic-level myelomeningocele compared with cystatin C estimates. All 3 cystatin C-based formulas were stable across lesion levels (P >.39) and highly correlated (ρ = 0.84-0.94); P30 exceeded 90% for every pair. We identified biases up to ±7.8 mL/min/1.73 m² and limits of agreement wider than ±40 mL/min/1.73 m².

Significant discrepancies were observed among creatinine-based equations, which consistently estimated kidney function to be higher at higher lesion levels (relative to cystatin C) and in myelomeningocele. Cystatin C-based estimates were stable regardless of lesion level. However, clinicians should be cautious about the interchangeability of cystatin C-based estimates.