AimRenal toxicity causes major morbidity following chemotherapy- abnormal iGFRs may be detected in up to 73.7% of patients.1 Creatinine is universally used as a biomarker to track fluctuating function and to calculate surrogate glomerular filtration rate (GFR) in the form of estimating equations.2 There is concern regarding the suitability of creatinine as a biomarker in this population, and it is proposed that cystatin C as a biomarker alone and also included in estimating equations may offer improved clinical suitability and accuracy.3MethodsIn this prospective, longitudinal study over a period of 18 months, 132 combined isotope GFR (iGFR), creatinine and cystatin C measurements were taken from 48 paediatric oncology patients at a Northern Children’s Hospital. Correlation and agreement analysis was performed for both individual biomarkers and estimating equations. Sensitivity data, along with ROC curve analysis was performed for all biomarkers and estimating equations. Data from three identified patients was isolated to examine individual patient variation over time.ResultsCreatinine identified only 1/32 patients with an abnormal iGFR (<90 ml/min/1.73 m2) compared to cystatin C which identified 12/32. Creatinine values and both estimating equations failed to change significantly over a period of declining iGFR though cystatin C did show a significant inverse increase (p<0.05). Bland Altman analysis for both the creatinine and combined equation showed poor agreement (mean difference -64 ml/min/1.3 m2 and -20 ml/min/1.73 m2 respectively). All biomarkers and equations showed poor sensitivity to detect an abnormal iGFR either below 70 ml/min/1.73 m2 or 90 ml/min/1.73 m2. A transformation factor applied to the equations significantly improved the sensitivity and clinical applicability of all equations. The data from three individual patients failed to reveal any significant intra-patient relationships.ConclusionData from this study cannot support the use of creatinine or cystatin C as a single biomarker to monitor renal function in children undergoing chemotherapy. Newer cystatin C and creatinine combined equations, whilst offering statistical superiority, do not offer the clinical superiority to replace iGFR or provide a tool for accurate dose calculations. A transformation factor can be applied to the results gained from the estimating equations to significantly improve the detection of abnormal iGFR, though work in other patient cohorts is needed to support this. Previous work also supported the use of a transformation factor, though application of their transformation factor to this current cohort failed to replicate the 100% sensitivity findings previously demonstrated4. Three patients were identified from the cohort and their paired iGFR and estimated GFR were monitored prospectively, over a period of approximately a year. Significant variation was observed between iGFR and eGFR at each time point for all three patients and therefore personalisation of GFR estimation from baseline iGFR and demographic data could not be proposed. This requires exploration in a larger cohort with the possible inclusion of additional baseline variables.ReferencesCRUK Survival trends over time in Children’s Cancers. 1.2015. https://www.cancerresearchuk.org/health-professional/cancer-statistics/childrens-cancers/survival#heading-Two Accessed 28th March 2019.NICE ( 2013) CG169 Acute kidney injury: Prevention, detection and management of acute kidney injury up to the point of renal replacement therapy.Barnfield, MC, Burniston, MT, Reid, U, et al. Cystatin C in assessment of glomerular filtration rate in children and young adults suffering from cancer. Nuclear Medicine Communications 2013;34:609–614.Dodgshun, AJ, Quinlan, C, Sullivan, MJ. Cystatin C based equation accurately estimates glomerular filtration rate in children with solid and central nervous system tumours: enough evidence to change practice? Pediatric Blood and Cancer 2016;63:1535–1538.
Assessing glomerular filtration rate (GFR) in critically ill patients with acute kidney injury - true GFR versus urinary creatinine clearance and estimating equations