cystatin c: better predictor of bad outcomes than creatinine

 another article found significant discordance between creatinine and cystatin C in assessing renal function,  finding that the latter correlated most strongly with both renal and non-renal adverse clinical outcomes (see ckd cystatin c better than creat AJKD2023 in dropbox or doi: 10.1053/ j.ajkd.2023.04.002)

 

Details:

-- 158,601 adults who were in the Stockholm Creatinine Measurements (SCREAM) project and had testing of creatinine and cystatin C on the same day during a healthcare encounter, from 2010-2018

    -- >400,000 outpatient tests were done for both creatinine and cystatin C levels

-- mean age 62, 48% female

-- mean eGFR by creatinine ( eGFR_cr) 80; mean eGFR by cystatin C (eGFR_cys ) 73

-- mean stage of renal insufficiency: 39% eGFR_cr  90+; 38% eGFR_cr 60-89; 12% eGFR_cr 45-59; 7.1% eGFR_cr 30-44; 3.2% eGFR_cr 15-29, 0.7% eGFR_cr <15

-- mean urinary albumin-to-creat ratio: 12 mg/g

-- comorbidities: hypertension 59%, diabetes 8%, coronary heart disease 16%, heart failure 11%, stroke 7.4%, copd 5.9%, rest <5%

-- meds: hypertension 55%, RAS inhibitors 38%, diuretics 28%, statins 24%

 

-- main measurement: % difference in eGFR_cys  minus eGFR_cr , expressed as eGFR_diff

-- main outcomes: kidney failure with replacement therapy (KFRT), acute kidney injury (AKI), atherosclerotic cardiovascular disease (ASCVD), heart failure, and all-cause death

 

-- median followup 4.5 years

 

Results:

-- during the followup period:

    -- all-cause deaths: 36,587 people (10,442 attributed to cardiovascular disease)

    -- ASCVD: 7,625

    -- heart failure: 10,159

    -- AKI: 5,648

    -- KFRT: 1,709

 

-- eGFR_cys  was lower than  eGFR_cr 65% of the time (ie, there was a negative eGFR_diff), with mean of -10% (+/- SD of 25) and absolute difference in eGFR of -7 (+/- SD of 19), with 32% having negative absolute eGFR_diff of >15

    -- the eGFR_diff values were more profound in those with higher  eGFR_cr values

    -- overall, patients with larger negative eGFR_diff were older, more often female, had higher  eGFR_cr and albuminuria, and had more comorbid medical conditions (including hypertension, diabetes, cardiovascular disease, and cancer)

-- dividing the group into quartiles of eGFR_diff:

    -- quartile 1 (eGFR_cys <<eGFR_cr with range -83 to -28); mean eGFR_cys  42, mean eGFR_cr 71

    -- quartile 2 (eGFR_cys <eGFR_cr with range -27 to -10); mean eGFR_cys  64, mean eGFR_cr 78

    -- quartile 3 (eGFR_cys appox equal to eGFR_cr with range -9 to +5); mean eGFR_cys  87, mean eGFR_cr 88

    -- quartile 4 (eGFR_cys >eGFR_cr with +6 to +133) ; mean eGFR_cys  97, mean eGFR_cr 81 

 

-- outcomes:

   

 

 

 

 

Commentary:

-- Sweden has been in the forefront of testing for cystatin C in the world, after the Swedish researcher Anders Grubb identified cystatin C as a filtration marker in 1985

-- in 2012, KDIGO (Kidney Disease: Improving Global Outcomes) recommended measuring cystatin C for either confirmatory testing of abnormal creatinine levels and in situations when creatinine is not accurate enough for clinical decision-making

-- creatinine levels increase (and estimated GFR decrease) with increased muscle mass, as well as with increased protein intake (which may increase muscle mass) but also with the acute ingestion of cooked meats (https://diabetesjournals.org/care/article/37/2/483/29467/Effect-of-a-Cooked-Meat-Meal-on-Serum-Creatinine ), fluid status, diurnal variation, and:

    --there are several meds reported in the medical literature that can increase serum creatinine levels though do not reflect renal injury: cimetidine, trimethoprim/sulfamethoxazole, fenofibrate, IV cephalosporins, salicylates, pyrimethamine, corticosteroids, flucytosine, vitamin D metabolites, and I suspect others….

-- cystatin C is a low-molecular weight protein that is filtered and not reabsorbed, and can be falsely elevated by chronic inflammation, obesity, smoking, glucocorticoid use, and hyper/hypothyroidism. Higher levels are also associated with risk of developing diabetes

-- despite the KDIGO recommendations and recommendation by the American Society of Nephrology in 2021 for the “increased, routine, and timely use of cystatin C”, cystatin C is underutilized in the US

 

-- This study was very impressive in its magnitude of patients followed with the large number of simultaneous measurements of eGFR by both creatinine and cystatin C and many prespecified outcomes over several years, with the following findings:

    -- a substantial discordance between the measured eGFRs by creatinine and cystatin C, with the predominant finding of lower values by cystatin C, especially in those with baseline higher eGFRs by creatinine measurement

        -- about one in four patients had an  eGFR_cys more than 27% lower than their eGFR_cr on simultaneous measurements

    -- a dramatic and consistent decline in bad outcomes from those with the most discordant findings (highest difference between cystatin C vs creatinine, quartile 1) down to those with higher creatinine than cystatin C (quartile 4)

        -- the hazard ratios reversing from absolute HRs of about 1.5-3 (!!) for the panoply of adverse effects (all highly statistically significant) in those in quartile 1 , down to 20-40% fewer adverse effects in quartile 4 (ie, relative protection in the group with higher eGFR_cys than eGFR_cr)

            -- ie,  this trend in outcomes seemed to be consistently related to the eGFR_cys  only, since those with lower  than eGFR_cr had fewer adverse events than those with equal eGFR_cr and eGFR_cys (quartile 3, the reference group for the HR analyses)

        --so, they found pretty decisively that the cystatin C calculated eGFR was much more predictive of bad outcomes than the creatinine based one

    -- and, this was true for both renal and non-renal outcomes!!

 

-- a post-hoc analysis of the PARADIGM-HF study, which overall found that sacubitril-valsartan (Entresto) was better than enalapril for heart failure with reduced ejection fraction (https://gmodestmedblogs.blogspot.com/2020/02/sacubitrilvalsartan-dec-hosp-and-cv_3.html), found that after 8 months of follow-up:

    -- 35.7% of patients had at least a 10-point lower eGFR by cystatin C than creatinine

    -- and that more negative eGFR_diff values were associated with:

        -- 18% increase in primary endpoint (composite of cardiovasc mortality or heart failure hospitalization), p=0.008

        -- 34% increase in cardiovascular mortality, p=0.001

        -- 39% increase in all-cause mortality, p<0.001

        -- 31% increase in worsening renal function, p=0.05

        -- ie, similar findings to the large Stockholm study above

-- other smaller studies have found that another way of measuring the difference between creatinine- vs cystatin C-based measurements, assessing the eGFR_cys / eGFR_cr ratio <0.8, was also associated with an array of bad clinical outcomes

-- it also may be relevant to us that cystatin C-based assessment of eGFR is not affected by a person’s race (https://www.nih.gov/news-events/news-releases/nih-supported-study-suggests-alternative-race-based-kidney-function-calculations), though creatinine-based ones have included race in their calculator

 

-- it remains unclear why there is such a difference in these eGFR measurements: was it issues specific to creatinine (muscle mass, meat eaten, meds…); issues related to cystatin C (especially chronic inflammation, likely associated with comorbidities ranging from diabetes/hypertension/heart disease/obesity to air pollution/depression/stress); perhaps issues related to different aspects of renal function that led to different eGFR values (eg selective glomerular hypofiltration); issues related to the non-GFR determinants that affect the eGFR_cr  or eGFR_cys ; or some combinations of these factors and perhaps others

 

One speculation is that cystatin C levels are better than creatinine for predicting adverse clinical outcomes because:

-- creatinine is lower in the aging population because of the higher prevalence of low muscle mass

-- cystatin C may be elevated because of the higher prevalence in the aging population of comorbidities that elevate it, especially those leading to chronic inflammation (which are most of the chronic diseases we see)

-- and these 2 effects (low creatinine levels and high cystatin C levels from non-renal causes) would lead to exaggerated differences of the estimated GFR that more truly affects both the renal and nonrenal (esp cardiovascular) clinical outcomes (ie, the lower eGFR_cys  may largely be a reflection of the underlying chronic conditions that lower its level and are themselves associated with both higher renal and nonrenal outcomes, and this marker of low eGFR_cys  perhaps largely points to the high risk of the patients for adverse outcomes)

 

-- there was a community-based study finding that estimated GFR (eGFR) by using either creatinine or cystatin C was not very accurate when compared to an actually measured GFR (https://gmodestmedblogs.blogspot.com/2022/07/egfr-not-such-great-estimate-of-renal.html ), which suggests a couple of things:

    -- we should be very circumspect about using the strict classifications of CKD when assessing renally-cleared meds: an eGFR that is hovering around 30 but dips lower does not necessarily mean that the patient has plunged from stage 3b to stage 4 renal failure and needs to have med doses changed right away (ie, it would be useful to have a pretty accurate eGFR for dosing meds that are renally cleared....)

    -- this study was only assessing the specific differences in eGFR measurements and the real measured GFR and did not measure differences in clinical outcomes. So, this study does not contradict the findings in the current study of worse outcomes with lower estimated GFR by cystatin C vs creatinine

-- I was not able to find other studies comparing the accuracy of both of these eGFR measurements vs measured GFR, though a study done in India did find that the cystatin C-based eGFR was more accurate: A Comparison of Serum Cystatin C and Creatinine with Glomerular Filtration Rate in Indian Patients with Chronic Kidney Disease - PMC (nih.gov)

  

Limitations:

-- though this was a huge study with impressive quality/quantity of data, it was still an observational one, which only allows for conclusions about associations between the eGFRs measured and clinical outcomes, and not causality

-- though these were “real-world” patients having their renal function assessed, there was likely bias in the participants: cystatin C measurements were largely confined to people where the creatinine was suspected to be inaccurate (no mention in the study of the actual reasons for ordering the various kidney function tests in different people)

    -- other studies, though much smaller, had differences between cystatin C and creatinine in the 8-16% range

-- no data on the many potential causes of inaccuracies of creatinine and cystatin C based measurements: muscle mass, BMI, a fuller array of the potential causes of chronic inflammation as well as actual measures of systemic inflammation, drugs that might affect the eGFRs, etc: therefore no ability to try to control for these potential confounders

-- this study was done in one country (Sweden), where there might have been different specific exposures that might affect kidney function and may limit generalizing the results to other areas of the world

 

so, what makes sense in clinical practice:

-- this study and several of the smaller ones reinforce that cystatin C is much more highly correlated with bad renal and extra-renal complications than creatinine

    -- however, i presume that there are general problems with measuring cystatin C, including cost and availability of the test, that might be an obstacle to its use

-- since creatinine is so easily accessible, i think the following makes sense:

    -- get simultaneous measures of creatinine and cystatin C levels

        -- in many elderly, especially frail ones, the cystatin C is really important to assess renal function (several of my older patients with eGFR_cr in the CKD stage 1-2 range actually have eGFR_cys in the CKD 4-5 range!!); and the opposite in muscular individuals.

        -- it was notable that the most discordant results in the Stockholm study were in those with lower creatinine levels, suggesting that whenever possible everyone who is being assessed for renal function have cystatin C measured (ie, not limiting the testing to those who one suspects the creatinine-based eGFR may be inaccurate)

    -- if there is discordance between, with the eGFR_cys being lower than eGFR_cr, then rely on the cystatin C-based calculation as the most predictive of both renal and extra-renal complications

        -- and, since those with low eGFR_cys are at high risk of heart disease, we clinicians should do more aggressive counseling and treatment to improve their cardiovascular risk, especially getting the LDL cholesterol very low, further reinforce the importance of stopping smoking, etc

 

geoff

-----------------------------------

If you would like to be on the regular email list for upcoming blogs, please contact me at gmodest@bidmc.harvard.edu

  

to get access to all of the blogs:  go to http://gmodestmedblogs.blogspot.com/ to see the blogs in reverse chronological order

  -- click on 3 parallel lines top left, if you want to see blogs by category, then click on "labels" and choose a category​

  -- or you can just click on the magnifying glass on top right, then type in a name in the search box and get all the blogs with that name in them

 

please feel free to circulate this to others. also, if you send me their emails (gmodest@bidmc.harvard.edu), i can add them to the list