Glomerular Filtration Rate Based on Cystatin C More Accurate for Predicting CVD, Mortality

A study published in JAMA Network Open found that estimated glomerular filtration rate based on creatinine (eGFRcr)—specifically, the range from 45 to 59 mL/min/1.73 m²—does not capture as many individuals at high risk for cardiovascular disease (CVD) and mortality compared with eGFR based on cystatin C (eGFRcys), which is more sensitive and specific for CVD and mortality risks in mild chronic kidney disease (CKD).

The Kidney Disease: Improving Global Outcomes guidelines recommend testing eGFRcys to confirm CKD if there are no other markers of kidney damage and when eGFRcr is less accurate. This study aimed to determine if the use of eGFRcys as a concordance test for CKD would improve the detection of high-risk CKD, with the hypothesis that there would be a higher risk of adverse outcomes in patients who had a confirmed diagnosis of CKD based on eGFRcys testing.

All participants in this study were from the UK Biobank cohort study. eGFR was calculated using the Chronic Kidney Disease Epidemiology Collaboration (CKD-EPI) equation without race coefficients. Albuminuria was defined using the urine albumin to creatinine ratio.

Reduced eGFR was calculated according to the Kidney Disease: Improving Global Outcomes classification system and whether eGFRcr was concordant with eGFRcys. CKD status was split into 4 groups: no CKD (eGFR ≥ 60 mL/min/1.73 m²); eGFRcr G3 (eGFRcr < 60 mL/min/1.73 m² and eGFRcys ≥ 60 mL/min/1.73 m²); eGFRcys G3 (eGFRcys < 60 mL/min/1.73 m² and eGFRcr ≥ 60 mL/min/1.73 m²); and both G3 (eGFRcr and eGFRcys < 60 mL/min/1.73 m²).

There were 428,402 participants in this study, with a median age of 57 (IQR, 50-63) years, and 55.4% were female. There were 9335 deaths in the 76,629 older participants aged 65 to 73 years, as well as 5205 CVD events and 46 kidney failure events. There were 14,776 deaths, 9328 CVD events, and 124 kidney failure events in the 351,773 participants younger than 65 years.

Higher values of eGFRcr were associated with lower values of eGFRcys, especially in older participants. eGFRcys was generally higher in younger participants at lower values of eGFRcr. The median difference between eGFRcys and eGFRcr increased from 4 mL/min/1.73 m² in the youngest age group to 14 mL/min/1.73 m² in the oldest age group (70-73 years).

A total of 9.2% of older participants and 2.4% of younger participants had discordant classifications of eGFR less than 60 mL/min/1.73 m². eGFRcr did not detect CKD in 6278 older participants who had an eGFRcys of less than 60 mL/min/1.73 m².

The probability of kidney failure in 10 years was low in both age groups (older, 0.06%; younger, 0.04%) and the 10-year probability of CVD and all-cause mortality was higher than the probability of kidney failure in all CKD groups. Participants with eGFRcys CKD G3 had a nearly doubled probability of CVD and all-cause mortality in older participants and a more than doubled probability in younger adults compared with participants without CKD. Participants with eGFRcr G3 did not have an increased risk of either outcome compared with participants without CKD.

Increasing hazards of CVD occurred when participants had an eGFRcr above or below 90 mL/min/1.73 m². Participants had a higher hazard of CVD and all-cause mortality when they had evidence of CKD from both eGFRcr and eGFRcys or by eGFRcys alone compared with participants with evidence of CKD from neither of eGFRcys or eGFRcr. Participants did not have higher hazards of CVD or death if they only had evidence of CKD G3 through eGFRcr.

Area under the receiver operating curves also demonstrated that eGFRcys was better at discriminating risk of fatal and nonfatal CVD and all-cause mortality in older and younger participants.

There were some limitations to this study. Single baseline measurements for creatinine, cystatin C, and albuminuria levels were used to define CKD. The maximum age of participants was 73 years, which makes generalizability to older age groups unknown. Participants in the UK Biobank did not have measured GFR levels, which means that the level of risk associated with eGFRcr or eGFRcys cannot be confirmed to be solely related to kidney function. All results assume that the participants have normal urine albumin concentration.

The UK Biobank includes primarily White participants, which can limit the generalizability to participants of other races. eGFR may have been overestimated in older individuals of European ancestry due to the use of the CKD-EPI 2021 equation, which has been shown to have bias in the past. The UK Biobank also has participants with a lower prevalence of chronic diseases than the general population.

The researchers concluded that using eGFRcr alone to categorize CKD failed to capture a proportion of participants who had a higher risk of CVD and mortality. The eGFRcys measure is more sensitive and can improve the specificity of a diagnosis of high-risk CKD when there are higher risks of CVD and death compared with kidney failure.

Reference

Lees JS, Rutherford E, Stevens KI, et al. Assessment of cystatin C level for risk stratification in adults with chronic kidney disease. JAMA Netw Open. 2022;5(10):e2228300. doi:10.1001/jamanetworkopen.2022.38300