Duty Biochemist Scenarios

Karen Smith
Queen Elizabeth Hospital
Birmingham
Scenario 1

45 yr male
Sample from haematology outpatients
No clinical details

Sodium 140 mmol/L (134-146 mmol/L)

Potassium 4.3 mmol/L (3.4-5.2 mmol/L)
Urea 6.1 mmol/L (3.4-7.6 mmol/L)
Creatinine 142 µmol/L* (60-126 µmol/L)
Bilirubin 75 µmol/L * (<22 µmol/L)
ALT 20 U/L (5-41 U/L)
ALP 67 U/L (40-130 U/L)
Albumin 38 g/L (34-51 g/L)
LDH Haemolysed (135-225 U/L)
Hb 92 g/L* (135-180 g/L)
What is the most likely diagnosis?

1. Gilberts syndrome
2. Anaemia
3. AKI
4. Haemolytic anaemia

Sodium 140 mmol/L (134-146 mmol/L)

Potassium 4.3 mmol/L (3.4-5.2 mmol/L)
Urea 6.1 mmol/L (3.4-7.6 mmol/L)
Creatinine 142 µmol/L (60-126 µmol/L)
Bilirubin 75 µmol/L (<22 µmol/L)
ALT 20 U/L (5-41 U/L)
ALP 67 U/L (40-130 U/L)
Albumin 38 g/L (34-51 g/L)
LDH Haemolysed (135-225 U/L)
Hb 92 g/L (135-180 g/L)
What is the most likely diagnosis?

1. Gilberts syndrome
Mildly raised bilirubin so a possibility
2. Anaemia
Yes - low Hb
3. AKI
Slightly raised creatinine but no previous results
4. Haemolytic anaemia
Combination of a low Hb and a mildly raised bilirubin is characteristic of
haemolytic anaemia
Whilst you are duty biochemist, a doctor from a tertiary
referral centre phones for the results. When you explain
that the LDH result is unavailable due to haemolysis, the
doctor explains that the patient has Paroxysmal Nocturnal
Haemoglobinuria (PNH) and therefore has intravascular
haemolysis.

He asks if you can give him the LDH result, as they need to
make treatment decisions on this result. He says that this is
what happened in other hospitals he has phoned for
results.
Do you give him the result?

1. Yes

2. No
 The haemolysis index cut-off used for LDH results is low as
LDH is found in red blood cells at high concentrations

 The haemolysis index cannot differentiate between intra-

vascular and extra-vascular haemolysis

 A raised LDH level may not necessarily be caused by the

patients condition but the value is used to guide treatment

 It is easier to handle these types of queries if a department-

wide decision is made on issues such as these
Scenario 2

Your laboratory is about to install new automated analytical

equipment which measures indices on every sample (haemolysis,
icterus, lipaemia). The analyser software automatically removes
any result that exceeds the index recommended by the
manufacturer.
This replaces the previous protocol of visual inspection of samples
for signs of icterus, haemolysis and lipaemia.

Your laboratory also runs an off-label automated assay for

angiotensin converting enzyme (ACE).

The ACE kit insert states the following;

‘Lipaemic, hemolytic or icteric sera cannot be used for

ACE activity determination’
What do you do?

1. Determine potential levels of interference from haemolysis,

icterus and lipaemia from published literature

2. Continue with visual inspection of ACE samples

3. Carry out an in house study of potential interferences to

obtain suitable indices
 Serum indices should be established in house where they are
not provided by the manufacturer.

 Verification of manufacturers serum indices should be

considered

 Be aware of manufacturer differences in serum indices

• Units
• Method of measurement
• Assessment materials used
Scenario 3

A clinician phones you when you are DB and asks about

thyroglobulin results from his patient who has thyroid cancer.

Jan 25 ug/L Diagnosis

May 21 ug/L Post-thyroidectomy
Nov 22 ug/L Post-radioiodine

He asks why the thyroglobulin is raised even though the

patient has undergone treatment
What do you suggest?

1. Carry out imaging studies to check for thyroid tissue

remnants

2. Measure thyroglobulin antibodies

3. Measure thyroglobulin by an alternative method

4. Check the assay performance and reanalyse the latest

sample
What do you suggest?

1. Carry out imaging studies to check for remnants of thyroid

tissue – may be useful if there is a clinical concern

2. Measure thyroglobulin antibodies – may be useful but not

necessarily in isolation

3. Measure thyroglobulin by an alternative method – most

useful investigation to look at assay interference

4. Check the assay performance and reanalyse the latest

sample – the trend in this patient has continued for nearly a
year so any assay problems should have been investigated
Further investigations

Imaging showed no thyroid tissue present

Patient clinically well
 Further blood sample sent
• Thyroglobulin (radioimmunoassay): 21 ug/L
• Thyroglobulin (automated IMA assay): <0.2ug/L
• Thyroglobulin antibodies: 895 IU/L (0-40)
 Thyroglobulin
immunoassay interference
is common issue

 Endogenous thyroglobulin
antibodies are usually the
cause (but not always)

 Compare Tg RIA and IMA

results to detect assay
interference (false positive
in RIA and false negative in
IMA)
Scenario 4

Whilst you are duty biochemist you see the following thyroid
function test results

TSH 1.02 mIU/L (0.3-4.5 mIU/L)

fT4 37.8 pmol/L (10-22 pmol/L)

There are no clinical details

Results from 4 months previously are;

TSH 0.95 mIU/L (0.3-4.5 mIU/L)

fT4 35.8 pmol/L (10-22 pmol/L)
What is the most likely cause of these results?

1. The patient is on thyroxine

2. Thyroiditis
3. The patient is on phenytoin
4. Assay interference
What is the most likely cause of these results?

1. The patient is on thyroxine - possible

2. Thyroiditis - possible
3. The patient is on phenytoin - possible
4. Assay interference - possible
You phone the GP to discuss the results

The patient is clinically euthyroid, has no neck swelling and is not

taking any drugs (including thyroxine)

What is the most likely cause of these results?

1. The patient is on thyroxine

2. Thyroiditis
3. The patient is on phenytoin
4. Assay interference
You investigate the possibility of assay interference by sending the
sample to another lab for analysis.

The results are;

TSH mIU/L fT4 pmol/L

Your lab (Roche) 1.02 (0.3-4.5) 37.8 (10-22)

Referral lab (Centaur) 0.93 (0.4-4.0) 17.0 (9-20)

What is the most likely cause of the abnormal results?

 TSH mIU/L fT4 pmol/L
Your lab (Roche) 1.02 (0.3-4.5) 37.8 (10-22)
Referral lab (Centaur) 0.93 (0.4-4.0) 17.0 (9-20)

What is the most likely cause of the abnormal results?

1. Autoantibodies to T4
2. Familial dysalbuminaemic hyperthyroxinaemia (FDH)
3. Heterophillic antibodies
4. Interference involving a specific assay component
What is the most likely cause of the abnormal results?

1. Autoantibodies to T4
Possible - antibodies to fT4 can act differently between
different assay types
2. Familial dysalbuminaemic hyperthyroxinaemia (FDH)
Possible – but most assays give a raised fT4 result
3. Heterophillic antibodies
Usually associated with TSH assay interference
4. Interference involving a specific assay component
Possible – the fT4 measured by the alternative method is
normal
Interference involving a specific assay component
Roche competitive immunoassay for fT4

Separation

Patient sample
and ruthenium
Addition of biotinylated-
labelled anti-T4
T4 and streptavidin Induction of
coated beads Chemiluminescence
Interference involving a specific assay component
Roche competitive immunoassay for fT4

Separation

Patient sample
and ruthenium
Addition of biotinylated-
labelled anti-T4
T4 and streptavidin Induction of
coated beads Chemiluminescence

 This patient’s sample contained anti-ruthenium antibodies

 These diminish the amount of signal causing a falsely elevated fT4 result
 The Roche fT4 and fT3 assays are inaccurate in these patients but other
manufacturers assays are unaffected
 TSH mIU/L fT4 pmol/L
Your lab (Roche) 1.02 (0.3-4.5) 37.8 (10-22)
Referral lab (Centaur) 0.93 (0.4-4.0) 65.2 (9-20)

What is the most likely cause of the abnormal results?

1. Autoantibodies to T4
2. Familial dysalbuminaemic hyperthyroxinaemia (FDH)
3. Heterophillic antibodies
4. Interference involving a specific assay component
What is the most likely cause of the abnormal results?

1. Autoantibodies to T4
Possible - antibodies to fT4 can act differently between
different assay types
2. Familial dysalbuminaemic hyperthyroxinaemia (FDH)
Possible – but most assays give a moderately raised fT4
result
3. Heterophillic antibodies
Usually associated with TSH assay interference
4. Interference involving a specific assay component
Possible – the fT4 measured by the alternative method is
also abnormal but both assays may demonstrate
interference
Interference involving autoantibodies to T4 (THAAb’s)

 fT4 (or fT3) levels can be unaffected, falsely elevated or

depressed depending on assay design and on the THAAbs
specificity of for the assay label.
 One step assays - THAAbs bind to assay label (labelled T4) so the
amount of label available for assay antibody binding is reduced
leading to falsely high results
 Two step assays – a wash step separates serum from label so any
interference should be largely eliminated
 In some assays THAAbs may bind antigen to a greater extent
than they do the assay label causing a falsely low result.
 One step-assay
Patient sera,
biotinylated-T4 and
streptavidin coated
beads

Two step-assay
Patient sera
incubated with anti-
T4 coated beads

separation
 One step-assay
Patient sera,
biotinylated-T4 and
streptavidin coated
beads

Two step-assay
Patient sera
incubated with anti-
T4 coated beads

separation
Method FT4 pmol/L TSH mU/L FT3 pmol/L
Centaur 55.3 (9-20) 1.10 (0.4-5.5) 5.5 (3.5-6.5)
Immulite 50.8 (10-25) 0.93 (0.3-5.5) 9.1 (2.5-6.5)
Tosoh ALA 1800 27.3 (10.6-21) 1.08 (0.4-4.0) 5.6 (3.2-5.9)
Roche E170 24.9 (10-22) 1.04 (0.3-5.5) 6.4 (3.1-6.8)
Equilbrium Dialysis 16.1 (10-36) (Nichols) -
Delfia 14.1 (9-20) 0.97 (0.4-4.0) 6.8 (3.0 -7.5)
Abbott Architect 9.0 (9-19) 0.93 (0.2-5.0) TT3 1.9 (0.9-2.4)nmol/L

Beckman DXI 800 7.3 (7-17) 0.96 (0.4-4.5) -

Soha A. Zouwail etal Red = one-step assay
Clinical Chemistry Green = one step (two
2008;54:927-928. incubation) assay
Blue = two-step assay

 Only equilibrium dialysis gives accurate values for free T4

independent of any interfering substances present
 Monitor patients clinically and by TSH alone
 TSH mIU/L fT4 pmol/L
Your lab (Roche) 1.02 (0.3-4.5) 37.8 (10-22)
Referral lab (Centaur) 0.93 (0.4-4.0) 35.1 (9-20)

What is the most likely cause of the abnormal results?

1. Autoantibodies to T4
2. Familial dysalbuminaemic hyperthyroxinaemia (FDH)
3. Heterophillic antibodies
4. Interference involving a specific assay component
What is the most likely cause of the abnormal results?

1. Autoantibodies to T4
Possible - antibodies to fT4 can cause a falsely high result in
a number of assays
2. Familial dysalbuminaemic hyperthyroxinaemia (FDH)
Possible –most assays give a moderately raised fT4 result
3. Heterophillic antibodies
Usually associated with TSH assay interference
4. Interference involving a specific assay component
Possible – the fT4 measured by the alternative method is
also abnormal but both assays may demonstrate
interference
Familial dysalbuminaemic hyperthyroxinaemia (FDH)

 FDH is a rare autodominant condition caused by mutations in

the serum albumin gene

 Can be confirmed with genetics or an FDH screening assay

 FDH is associated with increased total T4 concentrations and

normal physiological thyroid function

 The variant albumin has increased affinity (up to 60 fold) for

T4, so albumin is the major transport protein in FDH
 FT4 assay design assumes negligible binding of label to
albumin. However in FDH variant albumin exhibits significant
binding to labelled T4. So in FDH serum FT4 results can be
misleading.
One step methods - serum, label & antibody are all mixed
together. As variant albumin exhibits significant binding to
labelled T4, less label is available for assay binding. FT4 levels
will be falsely elevated.

Two step methods - serum components are separated from

the labelled analog T4 by the wash step. Abnormal binding of
labelled T4 should be avoided.

The method design & its constituents are also important –

some methods contain inhibitors of T4 binding to albumin,
in these FT4 would be elevated as TT4 is high in FDH.
Method FT4 pmol/L TSH mU/L FT3 pmol/L
Immulite 30.2 (10-24) 2.76 (0.4-4.0) -
Tosoh ALA 1800 25.7 (10.6-21) 3.23 (0.4-4.0) 5.0 (3.2-5.9)
Centaur 25.6 (11.5-22.7) 2.96 (0.4-5.55) 6.52 (3.5-6.5)
Roche E170 24.2 (10-22) 3.37 (0.3-4.5) 5.7 (3.1-6.8)
Beckman DXI 800 27.6 (7-17) 2.67 (0.4-4.5) 4.5 (3.5-6.5)
Abbott Architect 19.4 (9-19.1) 2.55 (0.35-4.94) 5.5 (2.6-5.7)
Delfia 16.8 (9-20) 2.74 (0.4-4.0) 6.6 (3.0-7.5)
Red = one-step assay
Green = one step (two
incubation) assay
Blue = two-step assay
Investigation of interference in thyroid function tests
 TFTs are ambiguous if the results are not consistent and/or they
do not agree with the clinical status of the patient.

 Common causes include analytical interference, drugs

(Amiodarone, Phenytoin), rheumatoid factor, sick euthyroid
syndrome or rarer clinical conditions such as FDH, TSHoma or
thyroid hormone resistance.

 If the cause is not apparent suggest exclude interference before

investigation of the rare clinical conditions

 Confirmation of an ambiguous TFT result by a different method

does not always rule out the presence of an interfering substance.

 The cause can often be hard to determine but specialised thyroid

screening methods available
Scenario 5

Whilst you are duty biochemist you see the following results.

59yr male
GP patient
No clinical details

Sodium 140 mmol/L (134-146 mmol/L)

Potassium 6.6 mmol/L (3.4-5.2 mmol/L)
Urea 5.1 mmol/L (3.4-7.6 mmol/L)
Creatinine 98 µmol/L (60-126 µmol/L)

What do you check?

1. Time of receipt and centrifugation
2. Haemolysis
3. Calcium and ALP
4. Full blood count
All of the above

1. Time of collection, receipt and centrifugation

Sample collected and analysed within 6 hours
2. Haemolysis
The sample was not haemolysed
3. Calcium and ALP
The Ca/ALP were normal ruling out EDTA contamination
4. Full blood count
The patient has raised platelets – 650 x109/L (150-450)
Sodium 140 mmol/L (134-146 mmol/L)
Potassium 6.6 mmol/L (3.4-5.2 mmol/L)
Urea 5.1 mmol/L (3.4-7.6 mmol/L)
Creatinine 98 µmol/L (60-126 µmol/L)
Platelets 650 x109/L (150-450 x109/L)

What do you do?

1. Remove the potassium result

2. Authorise with a comment regarding the raised platelets
3. Phone the result to the GP and suggest the test is
repeated in a plasma tube
Platelets and hyperkalaemia

 Potassium is released from platelets

during clotting hence a raised platelet
count can cause hyperkalaemia
 This effect is not seen in plasma
samples as clotting does not occur
 With normal platelet counts, plasma K
is ~0.35 mmol/L lower than serum K
 A difference of >0.4 mmol/L between
plasma and serum may indicate
psuedohyperkalaemia
 Caution - the difference between
serum and plasma K does not have a
strong correlation with platelet count
Detecting interference when duty biochemist

 Be vigilant for:
• Sets of results that are unusual (eg. TFTs)
• Results that do not fit with the clinical picture (eg. very low
creatinine in a renal patient)
• Results that do not fit physiologically (eg. high testosterone in
a female)
• Results that have changed significantly in a short time period

 Always recheck any suspicious results

 Discrepant results can be due to many different causes;

• Sample swaps
• Analytical errors
• Collection errors
• Assay interference
Any questions?