Klinisk Biokemi i Norden · 1 2020
| 9
research has benefitted from tight cooperation bet-
ween clinicians, clinical chemists and the diagnostic
industry, which has propelled the field forward. High-
sensitivity troponin assays were introduced into clini-
cal practice in 2009. Since then we got the rapid 0/1,
0/2, and 0/3 hour protocols for diagnosing NSTEMI,
still struggling with safety issues and striving to get
acknowledged in cardiology departments worldwide.
We got evidence that stably elevated troponin concen-
trations are one of the strongest predictors of future
major adverse cardiovascular events and death, in
whatever cohort or clinical situation you may choose
to investigate. We do not know why a stably elevated
troponin concentration is bad for you, but it is see-
mingly
not
related to the acute coronary syndrome.
We know that strenuous physical exercise will lead to
acute troponin release, but we do not know why, nor
if it is bad for you. Recently we have also learned that
fragmented forms of troponin molecules circulate in
certain conditions. The clinical utility of this, if any,
is currently unknown.
In Trondheim, we will meet several leading front-
line cardiology and clinical chemistry researchers
working with cardiac biomarkers. The congress
encompasses a keynote lecture from Fred Apple,
a legendary authority with more than 40 years of
experience in the field. He will give an overview of
the clinical impact of troponin analysis and how it
has changed our understanding of the heart. The
program also includes two parallel sections. One
is devoted to the interpretation of troponin results
in the setting of non-coronary disease and includes
several prominent speakers. Stefan Blankenberg
(University of Hamburg, Germany) will talk about
troponin measurements in the general population
as a way of screening for individuals with increased
cardiovascular risk. Torbjørn Omland (University of
Oslo, Norway) will describe how we may use cardiac
biomarkers to guide the treatment of cancer patients.
Furthermore, Steven Meex (University of Maast-
richt, The Netherlands) will discuss the utility of
measuring circulating troponin fragments and Stein
Ørn (University of Stavanger, Norway) will outline
the hitherto largest study for investigating exercise-
induced troponin release. A separate session will
focus on troponin in the more familiar setting of the
acute coronary syndrome. Different troponin-based
rule-out and rule-in algorithms will be presented by
Bertil Lindahl (University of Uppsala, Sweden) and
Camilla Hansen (Herlev Hospital, Denmark), while
Kristin M. Aakre (University of Bergen, Norway) will
focus on the analytical quality the laboratories must
provide if such algorithms are implemented in the
clinic. Finally, one of the leading researchers within
biomarker assessment of acute coronary syndrome,
Nick Mills (University of Edinburgh, United King-
dom), will present updated knowledge about patients
presenting with elevated troponin concentrations
who are not diagnosed with the acute coronary syn-
drome, a condition which since 2018 has been diag-
nosed as “myocardial injury”. These patients are at
high risk for future cardiovascular events, and large
studies are now undertaken to improve risk stratifica-
tion and improve outcomes in this group.
Analytical performance specifications
Arne Åsberg, Department of Clinical Chemistry,
St. Olavs hospital, Trondheim University Hospital,
Trondheim
Why do we need analytical performance specifi-
cations? We need those specifications in order to
make rational decisions when choosing analytical
instruments and methods, when choosing quality
control rules, when deciding on sample stability,
and when accepting reference intervals from other
laboratories – just to name some of the applications.
Analytical performance specifications are usually
specified as allowable bias and imprecision. Ideally,
the quality requirements should be based on the
effect of analytical performance on clinical outcomes
or, at least, the probabilities of certain outcomes. As
such data are lacking for almost all measurands, we
have to resort to the next best solution, i.e. to derive
allowable bias and imprecision from data on normal
biological variation. Traditionally, allowable bias has
been defined as one forth the normal total biological
coefficient of variation and allowable imprecision
as half the normal within-subject biological coef-
ficient of variation. Why? Because some honourable
experts have thought that this is good enough, not
by studying clinical outcomes, but by assumptions
and best judgement. There we stand today. However,
the session of analytical performance specifications
starring Theodorsson, Oosterhuis, Sandberg, and
Aarsand brings the hope of a better future. Certainly,
the speakers will give us some practical knowledge
to take home.
