https://orcid.org/0000-0003-3387-2123
WHAT IS CYSTATIN C?
Cystatin C has many similarities to creatinine. It is a small molecule (13 kDa) that is filtered unchanged at the glomerulus and is measurable in the serum. Serum cystatin C (as for creatinine) can therefore be used to estimate kidney function. Creatinine is cheap, widely available and measurement is standardized worldwide with the use of isotope dilution mass spectrometry–traceable laboratory techniques. Cystatin C has been proposed as a better, more accurate biomarker for estimating kidney function: it is released by all nucleated cells and is not influenced by age, gender, ethnicity or muscle mass. However, it is less widely available, is generally more expensive than creatinine (owing to the cost of reagents and lack of widespread use) and can be elevated in inflammatory states such as cancer, thyroid disease and corticosteroid use.
WHY USE CYSTATIN C TO ESTIMATE KIDNEY FUNCTION?
Historically, glomerular filtration rate (GFR) was measured using either exogenous (iohexol or chromium-51 ethylenediaminetetraacetic acid) clearance or endogenous markers, however, these markers are difficult to obtain, measurements are time-consuming and require expertise and there is substantial variability in the readings. Measuring endogenous markers (such as creatinine or cystatin C) in the urine over 24 h is similarly cumbersome and therefore impractical and inaccurate for widespread, repeated use.
Instead, GFR is estimated from serum biomarkers using one of many equations. Of these, the Chronic Kidney Disease Epidemiology Collaboration is thought to have the greatest precision and least bias, but others include the Modification of Diet in Renal Disease, full age spectrum and Berlin Initiative Study. For monitoring of changes in kidney function, estimated GFR (eGFR) is considered to be at least as good as measured GFR [1]. The eGFR can be calculated using creatinine alone (eGFRcr), cystatin C alone (eGFRcys) or a combination of creatinine and cystatin C (eGFRcr-cys) [2].
Under circumstances where it is desirable for eGFR to reflect renal excretory function as accurately as possible (purists may argue that this is desirable at all times), eGFRcr-cys is more closely aligned with measured kidney function than eGFRcr or eGFRcys alone across all levels of kidney function [2–4]. This may be because the non-GFR determinants (biases) of both creatinine and cystatin are independent and therefore can be attenuated in an equation that combines both biomarkers.
Chronic kidney disease (CKD) staging was generated as a simple means of planning follow-up and management, including complications associated with kidney disease [such as cardiovascular disease (CVD), anaemia and mineral and bone disorder], and of estimating the risk of end-stage kidney disease (ESKD). In the early stages, CKD is picked up more readily by eGFRcys or eGFRcr-cys [5]. Confirmatory testing with cystatin C–based eGFR is recommended for those with CKD Stage 3a (eGFR 45–60 mL/min/1.73 m2) [6], but we would recommend expanding these guidelines to those with eGFR 60–90 mL/min/1.73 m2 [5], particularly if they have additional risk factors for CVD or are at high risk of progressive CKD. For prediction of ESKD, both eGFRcys and eGFRcr-cys consistently outperform eGFRcr [5, 7].
HOW DOES CYSTATIN C IMPROVE PREDICTION OF CARDIOVASCULAR RISK?
Only a tiny portion of patients with CKD will progress to ESKD. Due to a combination of traditional atherosclerotic and CKD-specific risk factors, a substantial portion of patients with CKD will experience CVD (sometimes fatal) long before needing renal replacement therapy. This risk begins early in the disease process, increases as CKD progresses and includes atherosclerotic disease, heart failure, arrhythmia and sudden death [5, 7, 8]. The eGFRcys has a stronger, more linear relationship with cardiovascular outcome and mortality and is therefore able to detect an increased risk of CVD earlier than eGFRcr, and before clinically significant CKD becomes apparent. The eGFRcys detects an increased risk of CVD as soon as eGFRcys falls to <90 mL/min/1.73 m2, whereas eGFRcr does not consistently recognize this increased risk until eGFRcr falls to <∼75 mL/min/1.73 m2 [5, 7]. While it is likely that there are some non-atherosclerotic risk factors to blame, detecting increased cardiovascular risk earlier in the disease process may allow heightened awareness for patients and clinicians, implementation of lifestyle changes and early primary prevention strategies to attenuate the risk of CVD. Once CKD is established (especially when eGFR is <45 mL/min/1.73 m2), the added risk prediction provided by eGFRcys is less apparent [5].
It is likely that eGFRcys has a stronger relationship with cardiovascular risk than eGFRcr in part because of non-GFR determinants. Cystatin C is elevated in some inflammatory conditions with a slightly lower eGFRcys. These inflammatory conditions (themselves associated with cardiovascular risk) may be undiagnosed or are not captured routinely in risk prediction models. The lower eGFRcys can fulfil some helpful confounding in identifying those at higher risk.
IMPLICATIONS
The eGFRcys or eGFRcr-cys outperform eGFRcr for estimation of kidney function, diagnosis and staging of CKD and risk prediction of ESKD and CVD. Cystatin-based measures can also be safely used for drug dosing of potentially nephrotoxic medications [9].
The reluctance to use these markers in clinical practice likely relates to concerns over switching from cheap, available creatinine to a more expensive, less available biomarker, with relatively minor improvements in prediction of renal function, ESKD and CVD. The eGFR-C trial is designed to explore the accuracy, precision and cost-effectiveness of using eGFRcys to monitor CKD: results are expected later in 2020 [10].
While eGFRcys or eGFRcr-cys should not necessarily replace eGFRcr in widespread practice for routine monitoring of disease progression, earlier and more accurate detection of CKD and its associated risks could have a substantial effect on patient management and outcomes on a population scale.
SUMMARY
These differing eGFR measures could be used in the following ways to improve accuracy and patient outcomes:
eGFRcr-cys provides the most accurate estimation of GFR and could be used particularly when this accuracy is desirable, e.g. for drug dosing, consideration of organ donation or eligibility for clinical trials.
eGFRcys and eGFRcr-cys in anybody with an eGFRcr of 45–60 mL/min/1.73 m2 or 60–90 mL/min/1.73 m2 plus another cardiovascular risk factor to confirm diagnosis/staging of CKD. The lowest identified eGFR should be used for forward planning.
eGFRcys should be used alongside traditional cardiovascular prediction tools to enhance risk prediction and primary prevention of CVD.
eGFRcr for monitoring changes in kidney function, pending the outcome of cost–utility and cost-effectiveness trials.
CONFLICT OF INTEREST STATEMENT
This article has not been published previously, in whole or in part. There are no new data presented in this article. All results published previously have been referenced.
REFERENCES
Comments