![]()
24
| 1 | 2006
Klinisk Biokemi i Norden
Performance
Table 1 summarises the ability of S-Creatinine and
S-Cystatin C to detect reduced GFR expressed as
true and false positive (TP and FP), true and false
negative (TN and FN), specificity, sensitivity, (Pv
pos
)
and (Pv
neg
).
If only detection of GFR reduction below 60
mL/min is needed, creatinine performs as good as
Cystatin C. When the cut-off value between normal
and reduced GFR increases from 60 to 90 mL/min,
the number of false positives using creatinine
increases. Cystatin C performs better as it can be
seen in the Table 1.
Discussion
Many authors have proposed Cystatin C analysis as
a practical, cheap and safe replacement for crea-
tinine clearance as an indirect measure of kidney
function.
The problem with creatinine is its dependence on
sex, age, weight, body mass, disease and analytical
interference depending on analytical method. The
problem with creatinine clearance is especially
related to the difficulties in collecting 24-h urine
correctly.
Clearly a substitute for Creatinine evading all
these limitations and presenting a good estimation
of GFR would be welcomed.
Larsson (12) and others (8, 9, 11) have proposed
that Cystatin C is such a marker.
To support this claim an investigation is needed
showing not only that Cystatin C correlates better
to the true GFR than S-Creatinine, but also that
Cystatin C correlates better to the true GFR than
GFR calculated from S-Creatinine. Different algo-
rithms are proposed for adults, for children, for
special patients groups, e.g. Cockcroft and Gault
algorithm (3), Schwartz for children (5), Wright for
cancer patients (18), Levey (4), or the simplified
4-variables Levey’s formula (19, 20) recommended
by MDRD (4). If Cystatin C still will perform better
than GFR calculated from Creatinine according to
the optimal algorithm and will provide economical
neutrality, the case would be clear, and decision in
favour of Cystatin C easy.
Larsson has shown that the overall correlation of
Cystatin C to GFR is better than for S-Creatinine. If
however the results are grouped as recommended
by National Kidney Foundation (7) according to
kidney damage it can be seen that for severely,
moderately and slightly reduced renal function
Cystatin C does not perform significantly better
than S-Creatinine. However, Cystatin C correlates
better with GFR than S Creatinine when kidney
function is normal. The paper of Randers et al (15)
also supports this finding.
So, a clinician deciding what type of investiga-
tion – Creatinine, Creatinine clearance, Cystatin C
or GFR – to use, should first consider the purpose
of his GFR requisition. For monitoring treatment
Fig. 1.
Difference between GFR
calculated from S-Creatinine
(respectively Cystatin C) and
GFR measured.
mean diff (creat) = -2 ± 15 mL/min
mean diff (cyst) = -1 ± 10 mL/min
GFR measured, mL/min *1.73m
2
GFR calculated minus measured
(Fortsat fra side 21)
