All specialties have been advanced by

methods
to aid diagnosis, and monitor and assess
treat
ment. Investigation in endocrinology and
diabetes
remains centred on laboratory assays that
determine
the
concentration
of
hormones
and
metabolites
usually in blood. The first challenge is correct
col
lection; for some investigations, prior fasting
is
important.
Appropriate
conditions
or
preservatives
are mandatory (Box 4.1). In addition to
clinical
biochemistry
(also
called
chemical
pathology),
molecular genetics and cytogenetics are
routine
investigations to provide personalized genetic
diag
noses that predict the course of some
endocrine
disorders (e.g. multiple endocrine neoplasia;
see
Chapter 10).
Outside
of
the
laboratory,
clinical
investigation draws heavily upon expertise
in radiology and nuclear medicine. Some
investigations are highly specific (e.g. visual
fields for pituitary tumours or retinal screening
for diabetes) and these are covered in later
topic-specific chapters.

Laboratory assay platforms

Immunoassays
Hormones (and other metabolites) are most
com
monly measured by immunoassay, although
increas
ingly mass spectrometry is used (see
below).
Immunoassays, introduced in the 1960s, are
suffi
ciently sensitive, precise and
hormone
specific
for
routine
application
in
clinical
biochemistry.

Bioassays, which measure physiological

responses
induced by a stimulus, are near obsolete in
clinical
practice.
Immunoassay is a broad term for one of
two
different techniques: true immunoassay and
immu
nometric assay. Both forms are based on
the
hormone to be measured being antigenic and
bound
by specific antibodies to form an antibodyantigen
complex. Both forms of immunoassay also
employ
a label, historically a radioactive isotope [e.g.
iodine125 (I125)], but commonly now a fluorescent
tracer,
to generate a quantitative signal. Both assays
also
rely on comparison of the patient sample
with
known
concentrations
of
a
reference
compound.

Box 4.1 The specifics

collection are mandatory

of

sample

Containers for investigation of blood

Lithium heparin: most hormones
Fluoride oxalate: glucose
Di-potassium EDTA: tests requiring
isolation
Peptide hormones and catecholamines
tend to be less stable than other hormones
and need prompt delivery to the laboratory
on ice

DNA

Samples for which prior fasting may be required

Glucose
Lipids
Calcium
Hormones that may require 9
Cortisol
Testosterone

am

collection

Containers for investigation of urine

Acid: calcium, 5-hydroxyindoleacetic acid
(5-HIAA), catecholamines
No acid/simple preservative: urinary
cortisol

free

To set up a calibration or standard curve for

the
immunoassay, a constant amount of antibody
is
added to a series of tubes with increasing,
known
amounts of a reference preparation, in this
example
GH (Figure 4.1). This reaction is reversible with
the
antigen and antibody continuously associating
and
dissociating;
however,
after
incubation,
equilibrium
is reached when tubes with more GH generate
more
bound complex. Measurement of the amount
of
bound complex (e.g. in terms of fluorescence
or
radioactivity) can thus be related to the
quantity
of
GH that was originally added. This allows a
calibra
tion curve to be plotted, against which the
same
process can now deduce the GH concentration
in
a patient sample.

50 / Chapter 4: Investigations in endocrinology and diabetes

Zero GH
Zero GH

+2 GH

+4 GH

Anti-GH

+4 GH

GH

Labelled second anti-GH

Incubate

Incubate
Add excess labelled second anti-GH

+ excess

+ excess

Separate antibody-bound complex

Anti-GH

GH

Figure 4.1 The basics of immunoassay are shown

for growth hormone (GH; see text for details). For
clarity, in Figures 4.1-4.3 only small numbers of

Calibration curve
Tube 1

Tube 2

sample tube

hormone molecules and antibodies are shown; in

practice, numbers are in the order of 10 8-1013.

Tube
Count bound

Immunometric assays - the sandwich

assays
In the immunometric assay (shown
for
GH
in
Figure 4.2), a constant amount of
antibody
is
added
to each tube with increasing, known
amounts
of
reference preparation. After incubation,
the
amount
of GH bound to the antibody is detected
by
adding
an excess of a second labelled antibody
to
all
tubes.
The second antibody is directed against
a
different
antigenic site on GH from the first
antibody
to
form a triple complex sandwiching GH
between
the
two antibodies. Any unbound antibody is

Counts from

Tube 2
hGH in
sample
tube

removed,
leaving the amount of triple complex
to
be
deter
mined by quantifying the bound label
(e.g.
fluores
cence or radioactivity). This emission
is
plotted
for
increasing,
known
amounts
of
reference
compound
to generate a calibration curve
(Figure
4.2).
In
prac
tice, five to eight concentrations
of
hormone
standard are used to generate a
precise
calibration
curve, against which patient samples
can be inter

Standard GH

Figure 4.2 The basics of an immunometric assay for

growth hormone (GH; also see text). As in Figure 4.1, in
practice large numbers of molecules are present for each reagent
and the incubation of the first and second
antibodies is usually simultaneous. Because the hormone is bound
between the two antibodies in the triple complex
(
), this assay is sometimes referred to as a sandwich assay.
Separation of the complex from the excess
labelled second antibody is usually achieved physically, e.g. by

precipitation and centrifugation (the supernatant contains

the unbound antibody and is discarded). This leaves the
bound labelled second antibody to be
quantified by counting radioactivity or measuring
fluorescence. The low measurement from the zero tube,
Tube 1, and the higher value from Tube 2 are plotted on
the calibration curve. Tube 1 is not zero because of minor
non-specific antibody binding. The calibration line is also
curved, rather than straight, because of the reversible
nature of the interaction between antibodies and their
antigens. In practice, five to eight calibration points are
used to construct the curve.

Antibody-bound counts

fraction

Chapter 4: Investigations in endocrinology and diabetes / 51

polated. The immunometric assay is suitable only

Zero tube
when the hormone to be measured
permits
discrete
binding
of
two
+12
antibodies. This would not work for
small hormones such as thyroxine (T 4)
or
triiodothyronine (T3), for which the
competitivebinding immunoassay system must be
used.
Immunoassays - the competitive-binding
assays
In the competitive-binding immunoassay
(shown
for T4 in Figure 4.3), constant amounts of
antibody
and labelled antigen are added to each
tube.
A
zero
tube is set up that contains labelled T 4,
as
well
as
a
tube that also includes a known amount
of
unla
belled standard T4. Incubation allows
the
antigenantibody complex to form. Since the
zero
tube
contains twice as much labelled T4 as
antibody,
half
of the labelled hormone will be bound and
the
other
half will remain free (i.e. in excess). In
the
other
tube, unlabelled and labelled T 4
compete
for
the
limited opportunity to bind antibody.
The
total
antibody-bound T4 is separated (e.g. by
precipita
tion) and the label measured (e.g. by
fluorescence
or radioactivity). There will be less
signal from the
second tube because of competition from
the
unla
belled T4; the decrease will be a
function
of
the
amount of unlabelled T4 added, i.e.
the
signal

Anti-T4

+T4
+12

+12

Labelled T4

Unlabelled T4 standard

Incubate

+9
+9

Separate antibody-bound complex

3
3

Tube 1

Tube 2
Calibration curve
Tube 1

Counts from
sample tube
Tube 2

decrease
s as the
amount
of
unlabell
ed
T4

increase
s,

Count bound fraction

T4 in sample tube

Analytical methods linked to mass

spectrometry
In some situations, immunoassays are
unreliable
or unavailable, commonly because
antibodies
lack
sufficient specificity, or there are
difficulties
with
measurements at low concentrations
(e.g.
serum
testosterone in women). This leads to
differences
in
measurements across different assay
platforms
that
inhibit
the
development
of
internationally
agreed
standards for diagnosis and care. For
some
steroid
or peptide hormones, or metabolic
intermediaries,
mass spectrometry (MS) is becoming
increasing

Antibody-bound counts

allowing
the
construction
of
a
calibration
curve
(Figure 4.3). For clinical use, standard T 4
is
replaced
by the patient sample, with all other
assay
condi
tions
kept
the
same.
As
for
immunometric
assays,
a
five to eight point calibration curve
offers
sufficient
precision for patient samples to be
interpolated.

Standard T4

Figure 4.3 The basics of an immunoassay for

thyroxine (T4; also see text). As in Figure 4.1, in
practice large numbers of molecules are present for
each reagent. Under the conditions shown, the
competition between equal amounts of labelled and
unlabelled T4 in Tube 2 will be such that, on average,
50% of the antibody binding sites will be occupied
by labelled T4. Because of competition between
labelled and unlabelled hormone for a limited
amount of antibody, this type of immunoassay is
sometimes called a competitive-binding assay. After
removing unbound label (as in Figure 4.2 legend),
the fluorescent or radioactive bound fraction is
quantified and a calibration curve constructed. In
practice, five to eight calibration points are used to
construct the curve.