Introduction and Purpose

Non-ideal chromatograms are common

I
Insure consistency
i t throughout
th h t the
th department
d t t
Insure consistency throughout assay run
Help discover problems earlier
What Are We Looking At?
Brief explanation of the chromatograms used in this
presentation
For each analyte, there are 4 peaks
What Are We Looking At?

The first peak is the analyte we are looking for

What Are We Looking At?

The second peak is the internal standard we spiked in

the sample
What Are We Looking At?

Concentration of a sample is determined by comparing

the first peak’s abundance to the second peak’s
abundance
What Are We Looking At?

The last two peaks are used, via a ratio, to determine if

we have interferences in the first two peaks (aka
qualitative peaks)
Calibration Curve
A calibration curve is generated by taking standards of known concentrations and
comparing the area of the analyte compared to the area of the internal standard
This is done for several different concentrations and a best fit line through all the
points is generated
In the picture below, a sample that has a concentration of 50 (x-axis) would have
an the Analyte Area/IS Area (y-axis) of approximately 1.10 (the highest point on
the graph
If you looked at the peaks for this concentration 50 sample
sample, the analyte peak would
be slightly larger than the IS peak
Calibration Curve

We can use the cal curve to determine our concentration of our unknowns
Let’s say we have a sample that has a Analyte Area/IS Area of 0.40 (y-
axis) by looking at the corresponding x-axis
axis), x axis value
value, it means that we have
a concentration of approximately 17
The computer software does all this for you. This is to illustrate how
things work and where the concentrations are coming from
Integration Issues to Be Considered

Peak tailing
Integration Issues to Be Considered

Interfering peaks
Integration Issues to Be Considered

High background
Integration Issues to Be Considered

Inconsistent integration
Peak Tailing

Can occur on any assay

As long as all the controls are in and there is
no evidence of contamination, results can be
verified
The key to integrating peak tailing correctly is
to make sure all the peaks are integrated
consistently
Peak Tailing
Two different integration types
Not including tail
Including tail
Peak Tailing

Twenty MMA patients

plus controls were
integrated
Samples were first
integrated without
including the tail
Samples were then
integrated with the tail
included
The results were then
compared and
graphed
Peak Tailing: Take Home Message

Can be integrated accurately regardless if you include

the tail or not
Key to correctly integrating the run is to be consistent
throughout the entire run
Use your most honest integration
Future potential for additional standardization
Interfering Peaks

Can occur randomly or systematically

Even with perfect integration technique
technique, you cannot
always verify out the results
Samples
p must be accepted
p on a case-by-case
y basis
If interfering peaks are occurring chronically, it may
indicate that there is an underlying issue with the assay
and/or instrument which needs to be resolved
Interfering Peaks: General Guidelines

Peaks must show that they are ≥ 70% baseline

resolved, as illustrated by the following diagram
If in doubt, draw the lines, measure the distances in
millimeters, and perform the calculations
Make sure you have the right peak; use retention times
If in doubt, consult a member of management about
the integration
g
Interfering Peaks
Two types of integration
Valley to valley
Drop down to baseline
Interfering Peaks
So which is better?
Valley to valley
Drop down to baseline
Interfering Peaks
So which is better?
Valley to valley
Drop down to baseline
Interfering Peaks
So which is better?
Valley to valley
Drop down to baseline
Interfering Peaks
So which is better?
Valley to valley
Drop down to baseline
Interfering Peaks
So which is better?
Valley to valley
Drop down to baseline
Interfering Peaks
So which is better?
Valley to valley
Drop down to baseline
Interfering Peaks
So which is better?
Valley to Valley
Drop down to baseline
Interfering Peaks
So which is better?
Valley to Valley
Drop down to baseline
Interfering Peaks
Which is better?
When you have identical shaped peaks, dropping down to the
baseline is the more correct approach
pp
Interfering Peaks

What do you do when you have

two peaks that are not identical?
Interfering Peaks

What to do in a less than perfect world?

Interfering Peaks

What to do in a less than perfect world?

Interfering Peaks

What to do in a less than perfect world?

Case Study

How much of a difference does it really make?

In January 2006,
2006 the Adrenal Steroid assay underwent
a internal proficiency testing
The 17-Hydroxyprogesterone failed with results that
were 22.44
44 SD off of the mean
It was determined that the results were initially
incorrect because of the way the sample was
i t
integrated
t d
After rectifying the incorrect integration, the results fell
to within 0.27 SD of the mean
Case Study
The sample was initially integrated valley to valley
because there was a small interference right before the
peak of interest
p
Case Study
th sample
With the l being
b i integrated
i t t d valley
ll tot valley,
ll th quantitation
the tit ti
resulted in a value of 86 ng/dL
This was 2.44 SD off of the mean
Case Study
When the sample was re-extracted the next day,
day the interfering peak was
gone
The sample quantitated to a value of 107 ng/dL which is only 0.27 SD
from the mean
Case Study

Look at the peak of interest more closely

The black lines show an estimate of what the individual
peaks may look like on their own
Case Study

Shading helps better illustrate the peaks’ shape and

individual area
Case Study
While dropping down to the base line will pick up some
of the interfering peak, it is more accurate than leaving
off some of the peak of interest when you integrate
valley to valley
Case Study
While dropping down to the base line will pick up some
of the interfering peak, it is more accurate than leaving
off
ff some off the
th peakk off interest
i t t when
h you integrate
i t t
valley to valley
Case Study
After integrating the original sample with the interference by
dropping down to the baseline, the concentration of the sample
becomes 107 ng/dL which agrees with the re-extract
re extract
Case Study

How much of a difference does it really make?

Integrating by dropping down to the baseline proved to
be the more accurate integration method in this
example
Changing the integration from valley to valley to
dropping down to the baseline resulted in a 20%
change
h iin th
the results
lt
Had this been an actual pediatric patient, the results
may have been reported out with the wrong
interpretation and might have contributed to a
misdiagnosis of the child’s ailment
Interfering Peaks

What to do in a less than perfect world?

How would yyou accuratelyy integrate
g this p
peak?
Interfering Peaks

What to do in a less than perfect world?

How would yyou accuratelyy integrate
g this p
peak?
If ion ratios are out, there is nothing you can do
Interfering Peaks: Take Home Message

In cases where there is simple interference, it is best to

drop down to the base line, not integrate valley to
valley
How you integrate a peak can make a significant
difference in the reported value, interpretation, and
patent treatment
N t allll iinterfering
Not t f i peaks
k can b
be resolved
l d easily
il
In some cases, samples may need to be repeated and/or
reported
p as unable to q
quantitate due to an interfering
g
substance
Never compromise patient care
High Background

Can be caused by contamination of a reagent, a dirty

mass spec, and/or a less than optimal sample
These samples must be evaluated on a case-by-case
basis
It may mean that the sample or run needs to be re-
injected or re-extracted
Chronic high background needs to be brought to
attention because it can be an indication that
something needs to be corrected
High Background
T to
Try t use retention
t ti times
ti
If in doubt integrate valley to valley
Do not fiddle with the chromatogram to make it pass
Determine where the mean baseline is to use as guide
High Background
T to
Try t use retention
t ti times
ti
If in doubt integrate valley to valley
Do not fiddle with the chromatogram to make it pass
Determine where the mean baseline is to use as guide
High Background
Try to use retention times
If in doubt integrate valley to valley
Do not fiddle with the chromatogram to make it pass
Determine where the mean baseline is to use as guide
High Background
Try to use retention times
If in doubt integrate valley to valley
Do not fiddle with the chromatogram to make it pass
Determine where the mean baseline is to use as guide
High Background
Try to use retention times
If in doubt integrate valley to valley
Do not fiddle with the chromatogram to make it pass
Determine where the mean baseline is to use as guide
High Background

Try to use retention times

If in doubt integrate valley to valley
Do not fiddle with the chromatogram to make it pass
Determine where the mean baseline is to use as guide
High Background

Try to use retention times

If in doubt integrate valley to valley
Do not fiddle with the chromatogram to make it pass
Determine where the mean baseline is to use as guide
High Background

Try to use retention times

If in doubt integrate valley to valley
Do not fiddle with the chromatogram to make it pass
Determine where the mean baseline is to use as guide
High Background

Try to use retention times

If in doubt integrate valley to valley
Do not fiddle with the chromatogram to make it pass
Determine where the mean baseline is to use as guide
High Background

Try to use retention times

If in doubt integrate valley to valley
Do not fiddle with the chromatogram to make it pass
Determine where the mean baseline is to use as guide
High Background
Try to use retention times
If in doubt integrate valley to valley
Do not fiddle with the chromatogram to make it pass
Determine where the mean baseline is to use as guide
High Background: Take Home Message

Use relative retention time and mean baseline as a

guide
If in doubt integrate valley to valley
Do not change integration to make the sample pass
Talk to a member of management to see if a sample
can be accepted
The sample may have to be re-injected,
re injected, repeated,
and/or reported as “unable to quantitate due to
interfering substances”
Never compromise patient care
Inconsistent Integration

Can we adjust the integration of the calibrators and/or

controls to make the controls come in when there is
nothing obviously wrong with the integration or
chromatography?
Inconsistent Integration

NO!
Inconsistent Integration

Even a little bit?

Inconsistent Integration

NO!
Inconsistent Integration

What do we do when we have a run that has a

control that is only slightly out?
Inconsistent Integration

Follow the Westgard Rules for determining whether a run can still
pass
5% of the time you will have a control that is outside two standard
deviations
If the controls are consistently
y running
g on the high
g or low end of
the range it could mean there is a serious issue with the assay
that needs to be addressed quickly
If you let one run after another barely pass,
pass you could be adversely
affecting patient care
Show the run to a member of management
They may be able to make a judgment call on what the next course of
action should be
Example 1

How would you integrate these peaks?

Example 1

How would you integrate these peaks?

Example 2

How would you integrate these peaks?

Example 2

What was originally integrated

Example 2

How it should be integrated

Example 3

How would you integrate these peaks?

Example 3

How would you integrate these peaks?

Conclusions

Be consistent throughout your runs

Make sure all your peaks are integrated the same way as the
calibrators and controls
Pay particular attention to retention times, baseline,
and peak shape
Don’t hesitate to talk to someone in management
about a sample or run
Sometimes there is no way to accurately integrate a
sample
Never compromise patient care just to avoid having a
sample or batch of samples be repeated