how to achieve lower quantification limits · 2017. 8. 18. · 1 . how to achieve lower...
TRANSCRIPT
RUSSELL P. GRANT
LABORATORY CORPORATION OF AMERICA® HOLDINGS, BURLINGTON, NC, USA
1
HOW TO ACHIEVE LOWER QUANTIFICATION LIMITS: PART 1: FOUNDATIONS…for consideration
2
LC-MS/MS Principles of Analytical Measurement
Complexity reduction of sample – Isolation Ideally selective for analyte(s) – Sum of Complimentary techniques Ideally simple – balancing the 4 S’s Concentration of analyte(s) versus concentration of interferents External Calibration with internal standardization (IS) – Unique for MS (ECIDMS) Reproducible before Quantitative Recovery (for Isotope Dilution, add IS first)
Modified from Cooks, R.G. et al, J. Chem Edu, 59, 926-933 (1982) and Rick Yost (TAMS 2016)
0
20
40
60
80
100
120
1 2 3 4 5 6
Number of Techniques
Techniques 1) Extract 2) LC 3) MS 4) MS
Selectivity derived from Tandem “Hyphenated” Techniques
“The LC-MS/MS Experiment”
3
What defines the LLOQ? Bias and CV <20% ?
4
Accuracy (Bias): Metrology of Calibration, Lossless (corrected in assay process), Assay Selectivity, Purity of materials Precision (CV): Signal : Noise? Literally EVERYTHING involved in the assay and instrument (and all of the above plus repeatability over time)
Blank + IS
LLOQ
5
Pipetting precision (Contact dispense, Balance and EDTA plasma)
Mix the sample first Select the correct pipet and tip for
desired volume (10µL shown)
Know the imprecision of your pipetting system
(100µL pipet shown) and technique
Calibration slope: Solubility/Adsorptive Loss Trigger The “10 touch” experiment
6
Every surface (pipettes/glassware/plastics) for Stock and Calibrator manufacturing with the solute condition in consideration
Consider “concentration range” and ameliorate with chemistry and materials
Baseline (1 flask contact)
Pour over to fresh container
8x
Stressed – Compare to Baseline
Orotic Acid MeOH calibrator
y=0.66x
r2 = 1.000
Orotic Acid 0.1% NaOH calibrator
y=1.24x
r2 = 0.999
Combined
y=0.96x
r2 = 0.990
Addition to plates/tips of Calibrator Prior to IS addition
7
75%
80%
85%
90%
95%
100%
Baseline Stressed
Gluten peptide in PBS, pH 7.4, 0.5% BSA
-25%
-20%
-15%
-10%
-5%
0%
5%
1 3 5 7 9 11
Bia
s
Concentration (ng/mL)
Stressed Curves 2 and 3 vs stressed Curve 1 (3 randomized reps/curve plus samples 1 run)
75%
80%
85%
90%
95%
100%
Baseline Stressed
Gluten peptide in Tris, pH 8, 3% BSA
-25%
-20%
-15%
-10%
-5%
0%
5%
1 3 5 7 9
Bia
s
Concentration (ng/mL)
Stressed Curves 2 and 3 vs stressed Curve 1 (3 randomized reps/curve plus samples 1 run)
Even 15% Adsorptive loss = Calibration imprecision that IS cannot correct for See : Where Did My Analyte Go? – Coping with Poor Solubility and Non-Specific Binding, Catarina Horro Pita
Addition of IS in serial mode (4 dispenses per aspiration) Bias and Imprecision
8
Replicate/Tip 1 4 1 4 1 4
1 989299.1 927191.9 1049259.5 946906.7 1030562.8 911475.1
2 1029303.6 942528.5 1065516.4 991245.5 997552.9 899031.4
3 1061115.4 948176.9 1060470.1 989430.2 988456.5 884794.9
4 1060079.4 952630.4 1020440.0 966136.4 980805.5 906279.3
5 1051722.1 934064.9 1042351.4 961229.2 1016556.3 900131.1
6 1080978.6 954270.0 1060356.0 979481.4 988227.7 916055.2
7 998548.0 956196.8 Bias 1012357.3 985305.3 Bias 1007473.7 961248.3 Bias Average S.D. CV (%)
8 1026631.9 941873.8 Step 4 to 1 1077069.3 921012.7 Step 4 to 1 974348.4 928087.2 Step 4 to 1 984880.5 53348
5.41 Mean 1037209.8 944616.6
-8.9% 1048477.5 967593.4
-7.7% 997998.0 913387.8
-8.5% S.D. 32038.8 10219.2 22429.7 24294.7 19034.7 23199.1
CV (%) 3.1 1.1 2.1 2.5 1.9 2.5
Replicate/Tip 1 4 1 4 1 4
1 400675.1 390357.2 395417.9 394114.1 395814.7 402188.3
2 416891.7 397422.9 398281.4 395746.3 396414.7 404584.2
3 403699.8 392128.0 402128.5 395993.3 408201.3 402101.2
4 403017.0 394389.2 396687.1 397949.4 406706.0 393925.7
5 408418.1 397334.9 405532.6 398622.3 409820.6 401478.8
6 399635.7 392591.3 404128.7 399847.5 404494.2 399721.2
7 401346.2 393197.3 Bias 403289.1 402496.8 Bias 402245.8 398567.5 Bias Average S.D. CV (%)
8 404320.0 392820.4 Step 4 to 1 399895.9 401725.6 Step 4 to 1 401544.8 389046.8 Step 4 to 1 399936.6 5443.9
1.36 Mean 404750.5 393780.1
-2.7% 400670.2 398311.9
-0.6% 403155.2 398951.7
-1.0% S.D. 5599.9 2489.9 3671.2 2960.0 5169.0 5107.6
CV (%) 1.4 0.6 0.9 0.7 1.3 1.3
Replicate/Tip 1 4 1 4 1 4
1 392787.7 388537.1 393217.6 385940.9 382757.2 389361.9
2 400111.2 389610.5 389785.9 384788.7 384557.5 382013.7
3 394277.8 390411.6 388066.6 385958.7 389315.3 380779.6
4 397538.3 392501.1 392502.8 388224.3 380559.2 381812.1
5 387718.3 388289.5 389380.7 391244.4 382829.8 381799.5
6 396628.3 388850.7 388969.9 392289.7 383727.3 377978.7
7 388807.5 391158.5 Bias 383562.6 378048.8 Bias 384367.2 389546.0 Bias Average S.D. CV (%)
8 399712.6 385670.3 Step 4 to 1 388177.1 379897.1 Step 4 to 1 387594.3 390233.7 Step 4 to 1 387956.2 5202.7
1.34 Mean 394697.7 389378.7
-1.3% 389207.9 385799.1
-0.9% 384463.5 384190.7
-0.1% S.D. 4682.5 2068.9 2966.2 4980.4 2796.1 4755.1
CV (%) 1.2 0.5 0.8 1.3 0.7 1.2
IS Peak
Area
(counts) Including all data points
IS Peak
Area
(counts) Including all data points
3X Pre-wet
Addition Step Addition Step Addition Step
IS Peak
Area
(counts) Including all data points
1X Pre-wet
Addition Step Addition Step Addition Step
Baseline
Addition Step Addition Step Addition Step
1st vs 4th Delivery 8-tip liquid handler No “prewetting”
1X “Pre-wetted”
3X “Pre-wetted”
Pre-condition tips, Automation tips ≠ Manual pipet tips
Internal Standard addition – its not just about weight
9
Pre-wet tips (x3) with IS solution Single dispense D4-Cortisol IS includes 0.1% BSA
CV = 8.6%
8-tip liquid handler IS addition using same 8 tips
Use reagent additives to reduce adsorption, IS precision enables outlier detection
Elucidating Adsorptive Post IS
10
A
B
C
1 12
A1: 1 tip 1 well
B10: 1 tip transfer all sample over 10 wells
C1: 10 fresh tips aspirate and dispense into 1 well
B10 vs A1 adsorption to wells, C1 versus A1 - adsorption to tips
10x
Tip + Well Adsorption: Pipet sample into wells A1, B1 and C1
Consider every step and ameliorate with chemistry and materials
0%
20%
40%
60%
80%
100%
120%
unstressed (A1) 10 tips (B10) 10 wells (C1) 10 tips + 10 wells
Rel
ativ
e R
eco
very
H2O w/ Detergent 50:50 H2O:MeOH
Precision and “Maximum batch size” – Influence of Time
11 Two 96-well plates: Blanks and Calibrators, QC’s, Sample Pool
Plate 1 Plate 2
"Linear" Regression ("1 / x" weighting): y = 0.0314 x + 0.00358 (r = 0.9990)
10 20 30 40 50 60 70 80 90 100
Analyte Conc. / IS Conc.
0.0
0.5
1.0
1.5
2.0
2.5
3.0
3.3
Analy
te Area / IS
Area
Concentration Target (ng/mL)
Back-Fit Accuracy (%) Curve 1 Curve 2 Curve 1 re-inject
0.5 113.7 111.4 112.8 1 90.3 88.3 93.6 5 95.0 96.7 94.0
10 99.3 96.7 99.1 50 102.8 98.7 94.2
100 99.0 101.2 103.3
Which pipetting technique to use (8 or 96-tip)?
12
-30%
-20%
-10%
0%
10%
20%
30%
40%
50%
60%
0 50 100 150 200
-30%
-20%
-10%
0%
10%
20%
30%
40%
50%
60%
0 50 100 150 200
-30%
-20%
-10%
0%
10%
20%
30%
40%
50%
60%
0 50 100 150 200 250 300
-30%
-20%
-10%
0%
10%
20%
30%
40%
50%
60%
0 50 100 150 200
Manual CV = 5%
8/96/8 CV = 12%
8/96/96 CV = 8%
8/8/8 CV = 3%
Bia
s (%
)
Index
1: Add Sample 2: Add Generation buffer 3: Add IS (after generation)
Let repeatability of a pool guide automation technique
13
SLE precision of elution volume
0
0.02
0.04
0.06
0.08
0.1
0.12
0.14
0 100 200 300
Pre
ssu
re (p
si)
Time (sec)
0
5
10
15
20
25
30
35
40
0 200 400 600
Pre
ssu
re (p
si)
Time (sec)
LOAD: PRE-MIXED SAMPLE + IS (5 SEC 0.12PSI, HOLD 5 MIN)
ADD SOLVENT AND ELUTE (5 SEC 0.12PSI, HOLD 10 MIN, 30 SEC 36PSI)
-150%
-100%
-50%
0%
50%
100%
150%
0 10 20 30 40 50 60 70 80 90 100
Bia
s to
Glo
bal
Mea
n
Injection Order
No Pulse: IS Peak Area Bias to Global Mean
-150%
-100%
-50%
0%
50%
100%
150%
0 10 20 30 40 50 60 70 80 90 100
Bia
s to
Glo
bal
Mea
n
Injection Order
Pulse: IS Peak Area Bias to Global Mean
Multi-step SPE for Reverse T3: Very Complicated
14
Step Solvent Volume (mL) Function
1 CH3OH 1 Clean 96-well plate
2 H20 1 Pre-condition plate
3 Sample 0.4 Load 1:1 diluted sample containing IS
4 5% NH4OH:CH3CN 1:1 1 Elute soluble acids, proteins and neutrals
5 5% NH4OH 1 Elute aqueous soluble acids, proteins and neutrals
6 CH3OH 1 Elute acids and neutrals
7 C3H7OH 1 Elute lipids and polar analytes
8 2% HCOOH in CH2Cl2 1 Elute lipids and polar analytes
9 5% HCOOH in CH3OH 1 Elute anionic moieties
10 Evaporate 40 C, 45 minutes, reconstitute, seal plate, mix, centrifuge and inject
50 10
0
15
0
5
4
3
2
1Pre
ssure
(p
si)
Time (s)
SPE Plate Condition and Loading
50 10
0
15
0
5
4
3
2
1Pre
ssure
(p
si)
Time (s)
SPE Plate Washing and Eluting
Precision of technique is critical, particularly in complicated workflows
Using duplicate single plates – Curve and identical distributed samples on each plate (Note Percent Bias Scale)
15
Manual versus Automated Deming Slope 1.002
r 0.9994 Bias (%) -0.700
IS Precision (%) 16% vs 7%
ALDIII Left versus Right Deming Slope 0.961
r 0.9995 Bias (%) -1.345
IS Precision (%) 6% vs 8%
Use automation and perform triplicate runs for method comparison (reduce influence of single batch variance on comparative slope bias)
16
Part 1 summary..what we learned
You must know and improve the imprecision of pipetting techniques
Control of adsorptive losses is critical (concentration/solute/solvent/surface dependent)
Bias = imprecision over time
Internal standard response and single pool imprecision are go to tools
RUSSELL P. GRANT
LABORATORY CORPORATION OF AMERICA® HOLDINGS, BURLINGTON, NC, USA
17
HOW TO ACHIEVE LOWER QUANTIFICATION LIMITS: PART 2: FORMULATION...of an assay and system
18
Cleanliness of System
0.5 1.0 1.5 2.0
Time, min
8000
1.6e4
2.4e4
3.2e4
4.0e4
4.8e4
Inte
nsi
ty, c
ps
Blank
Blank after ”Top Off” Solvents
LLOQ, S:N = 54
Pre-injection Hold (100% A)
Noise
“Absence” of signal and statistics required to define LLOQ, not S:N
19
More “sensitivity” = More interferences
+0.02 ng/mL
Human Serum “Tg(-)/TgAb(-)”
Chicken Serum 99% HSA (60 mg/mL in PBS)
96% HSA (60 mg/mL in PBS)
Elution Control
LabCorp Surrogate Matrix
SigMatrix (20 mg/mL rHSA in PBS)
+0.04 ng/mL +0.08 ng/mL
Final µLC-MS/MS Surrogate Matrix
See: How Low Can You Go? Using Flow to Measure Zero, Chris Shuford
20
Precision of Recovery and Volumes (Organic Precipitant to Sample Ratio’s)
2mL Pooled Serum + 3:1, 2:1, 1:1 Acetonitrile or Methanol Mix 5 minutes, Centrifuge at 2500g for 10 minutes
Acetonitrile - smaller floccule, different “texture”, less precipitation than Methanol? Floccule size 3:1 > 2:1 > 1:1 – Subtle, precision of supernatant removal?
Turbidity 1:1 >2:1 > 3:1
Polson et al, Optimization of protein precipitation based upon effectiveness of protein removal and ionization effect in liquid chromatography-tandem mass spectrometry, J Chromatogr B Analyt Technol Biomed Life Sci. 2003 Mar 5;785(2):263-75.
Acetonitrile 3:1 2:1 1:1
Methanol 3:1 2:1 1:1
21
Post Column Infusion of Extracts to define LC needs
Acetonitrile PPT (10µL injected) 5 min Gradient 0 – 100% MeOH
50x2.1mm, 5µm XDB-C18 1mL/min Neat, 3:1, 2:1, 1:1
0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0 Time, min
0.0
1.0e5
2.0e5
3.0e5
4.0e5
5.0e5
6.0e5
7.0e5
8.0e5
9.0e5
1.0e6
1.1e6
1.2e6
1.3e6
1.4e6 1.5e6
Methanol PPT (10µL injected) 5 min Gradient 0 – 100% MeOH
50x2.1mm, 5µm XDB-C18 1mL/min Neat, 3:1, 2:1, 1:1
0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0 Time, min
0.0
1.0e5
2.0e5
3.0e5
4.0e5
5.0e5
6.0e5
7.0e5
8.0e5
9.0e5
1.0e6
1.1e6
1.2e6
1.3e6
1.4e6 1.4e6
0.0 1.0 2.0 3.0 4.0 Time, min
4.0e5
8.0e5
1.2e6
Inte
nsity, cp
s
1:1 ACN:H2O Calibrator Matrix Serum Sample Serum Separator Tube
22
Reducing dilution in Protein Precipitation
90:10 ACN:H2O +(NH4)2SO4 +ZnSO4 90:10 MeOH:H2O +(NH4)2SO4 +ZnSO4
Solvent + Saturated Salt (90:10 aqueous) at 1:1 PPT ratio
70% Perchloric Acid at various Acid:Sample PPT ratios
3:1 2:1 1:1 0.5:1 0.25:1 0.1:1 0.05:1
23
Centrifugal Force (not RPM) – Higher and longer is better
Database for centrifuge conversions: http://www.endmemo.com/bio/grpm.php
G = (1.118 x 10-5) x r x (RPM)2
G = RCF (relative centrifugal force) r = rotational radius (centimeter, cm) RPM = evolutions per minute
500 1000 1500 2000 2500g Acetonitrile PPT 3:1
Blank 3:1 2:1 1:1 Methanol PPT, 10,000g, 24h @10C
96-well plates Centrifuge at ~3600g *
24
Control your LC system dead volume
40µL injection into a 100µL Loop Analyte Area CV = 29%
IS Area CV = 21% Ratio CV = 13%
0.2 0.4 0.6 0.8 Time, min
2.00e4
4.00e4
6.00e4
8.00e4
1.00e5
Inten
sity, cps
0.2 0.4 0.6 0.8 Time, min
2.00e4
4.00e4
6.00e4
8.00e4
1.00e5
Inten
sity, cps
0.2 0.4 0.6 0.8 Time, min
2.00e4
4.00e4
6.00e4
8.00e4
1.00e5
Inten
sity, cps
40µL injection into a 20µL Loop Analyte Area CV = 12%
IS Area CV = 12% Ratio CV = 5%
Post-column “Orange versus Red peak” Analyte Area CV = 74%
IS Area CV = 71% Ratio CV = 40%
25
Why Isocratic LC eventually fails you
MeOH PPT, Isocratic MeOH:H2O
Early QC
25OH Vitamin D2
Later QC
25OH Vitamin D2
+ Gradient
+ Clean cycle
25OH Vitamin D2
Retinol D4 Retinol
0.05 0.15 0.25 Time, min
5.0e4
1.0e5
1.5e5
2.0e5
Inten
sity, cps
0.35
1.0e5
2.0e5
3.0e5
Inten
sity, cps
0.05 0.15 0.25 Time, min
0.35
5.0e4
1.0e5
1.5e5
2.0e5
2.5e5
Inten
sity, cps
0.05 0.15 0.25 Time, min
0.35
5.0e4
1.0e5
1.5e5
2.0e5
Inten
sity, cps
0.05 0.15 0.25 Time, min
0.35
Retinol D4 Retinol
IPA PPT, Gradient ACN:H2O
LLOQ injection 1st v 2nd curve
Bias ~90% + MeOH wash
Bias <7%, CV< 5%
Use gradients, better prep/LC and elutropic wash for “reproducible integration” See: Pragmatic HILIC for the Clinic by Brian Rappold
26
Importance of maintenance
Vitamin B6 (PLP)
0.10 0.20 0.30 Time, min
1.0e4
2.0e4
3.0e4
4.0e4
Inten
sity, cps
0.10 0.20 0.30 Time, min
1.0e5
2.0e5
3.0e5
Inten
sity, cps
D3-Vitamin B6 (PLP)
0.10 0.20 0.30 Time, min
1.0e4
2.0e4
3.0e4
Inten
sity, cps
1.0e5
2.0e5
3.0e5
Inten
sity, cps
0.10 0.20 0.30 Time, min
Vitamin B6 (PLP) D3-Vitamin B6 (PLP)
Ion sources are an electrochemical cell
System operates in Negative ion mode
Mean Bias = 4.4%
CV = 27.7%
New emitter/source clean
Mean Bias = 3.2% CV = 10.9%
Learn to interpret Noise around and superimposed on your analyte peak
27
Solvent (distillation) chemistry
0
10000
20000
30000
40000
50000
60000
70000
no DMSO 5% DMSO
Pe
ak
Are
a (c
ou
nts
)
2.1-fold increase p = 2 e-6
Alternating injections for A peptide (Load/Elute/Load/Elute…N=8)
28
0.1 0.2 0.3 0.4 Time, min 0.0
4000
8000
1.2e4 Inten
sity, cps
0.1 0.2 0.3 0.4 Time, min 0.0
8000
1.6e4
2.4e4
3.2e4
Inten
sity, cps
6 Fluphenazine transitions (10ms, 3 for QN and QL), product ions offset by 0.01amu
0.10 0.20 0.30 Time, min
500
1000
1500
2000
Inten
sity, cps
0.10 0.20 0.30 Time, min
400
800
1200
Inten
sity, cps
Norfentanyl 200pg/mL 233.20→84.0
Norfentanyl 200pg/mL 233.21→84.0
Norfentanyl 200pg/mL Transition Sum
0.10 0.20 0.30 Time, min
150
350
550
750
950
Inten
sity, cps
S:N 24.7 S:N 10.2 S:N 25.4
CV: 28.89% Bias: -5.60%
CV: 12.97% Bias: 18.55%
CV: 9.63% Bias: 6.50%
Transition summing: Discovered by accident in 2003
Quite simply you should do this...it literally feels like cheating
Single (30ms)
Summed (3*10mS)
S:N 16 - 40 50Bias (%) -10.0 1.2CV (%) 16.4 4.5
29
Scheduled MRM to solve noisy peaks from summation and ensure appropriate peak sampling, plus signal averaging
0.10 0.20 0.30 0.40 Time, min 0.0
1.5e4
3.0e4
4.5e4
6.0e4
Inten
sity, cps
SRM 24 transitions, 12 points/peak
0.10 0.20 0.30 0.40 Time, min 0.0
1.0e4
2.0e4
3.0e4 Inten
sity, cps
Scheduled MRM 24 transitions, 12 points/peak
SRM 2-5mS per transition
9.7% CV
Scheduled MRM 11 – 55 mS per transition,
2.4% CV
IS Peak Area Precision
IS R
esp
on
se
9-OH Risperidone
Chlorpromazine
Fluphenazine
Haloperidol
Methylphenidate
Risperidone
30
Peak sampling, Noise and Scheduled MRM: Impact on transition ratio precision
Selected Reaction Monitoring Scheduled MRM
Dwell (mS) Ion Ratio CV (%) Dwell (mS) Ion Ratio CV (%)
Methylphenidate 2 – 5 0.44 21.6 11 – 31 0.44 5.2
9-OH Risperidone 2 – 5 0.74 18.5 11 – 25 0.82 5.1
Risperidone 2 – 5 0.52 11.1 11 – 25 0.51 6.2 Haloperidol 2 – 5 2.07 6.6 25 – 55 2.07 3.0
Chlorpromazine 2 – 5 1.32 7.2 13 0.97 13.5* Fluphenazine 2 – 5 0.81 7.9 13 0.85 9.9
Short dwell times for peak sampling = Noisy data Scheduled MRM increases dwell time per transition = Signal averaging
Chl: 319/86, IS 322/89 Chl: 319/246, IS: 322/89
Analyte Conc. / IS Conc. 0
8
16
24
32
An
alyt
e A
rea
/ IS
Are
a
60
70
80
90
100
110
120
130
140
Chl: 319/86, IS 322/89 Chl: 319/246, IS: 322/89
Acc
ura
cy (%
) Injection Number
86 58
100 200 300 m/z, Da 0.0
4.0e7
8.0e7
1.2e8
1.6e8 Inte
nsity, cp
s
58
319 86
246 274
274
246
* Differential Fragmentation over time for Chlorpromazine
31
Modifying Quadrupole Resolution (Q1 and Q3) ..in the absence and presence of noise
“High” Resolution (0.5 amu, FWHM)
0.1 0.2 0.3 0.4 Time, min
500
1000
1500
2000
2500
Inte
nsi
ty, c
ps
Unit Resolution (0.7 amu, FWHM)
0.1 0.2 0.3 0.4 Time, min
500
1000
1500
2000
2500 In
ten
sity
, cp
s
0%
5%
10%
15%
20%
25%
30%
35%
40%
0.5 / 0.5 0.5 / 0.7 0.7 / 0.5 0.7 / 0.7 0.7 / 1.0 1.0 / 0.7 1.0 / 1.0
A:I
S (C
V)
Resolution, FWHM (Q1/Q3)
“Low” Resolution (1.0 amu, FWHM)
0.1 0.2 0.3 0.4 Time, min
500
1000
1500
2000
2500
Inte
nsi
ty, c
ps
0%
5%
10%
15%
20%
25%
30%
0.5 / 0.5 0.5 / 0.7 0.7 / 0.5 0.7 / 0.7 0.7 / 1.0 1.0 / 0.7 1.0 / 1.0A
:IS (C
V)
Resolution, FWHM (Q1/Q3)
Analyte to IS Precision
0.1 0.2 0.3 0.4 Time, min
1.0e4
2.0e4
3.0e4
4.0e4
Inten
sity, cps
0.5 0.1 0.2 0.3 Time, min
1.0e4
2.0e4
3.0e4
4.0e4
Inten
sity, cps
0.1 0.2 0.3 0.4 Time, min
1.0e4
2.0e4
3.0e4
4.0e4
Inten
sity, cps
0.5
Absence of noise = More “signal” Presence of noise (biological and/or chemical) use caution
32
Part 2 summary…what we learned
You must have a “clean system and matrix blank”
Calculation of signal:noise is somewhat irrelevant (can we agree?)
Sample preparation should enable precision of aspiration (volume/centrifugation)
Sample preparation ideally be non-dilutionary and must resolve matrix effects
LC system dead-volume must be minimized and use gradients
Maintain instrument source/optic cleanliness (plus use a bypass valve)
Ion source solvent chemistry is a key variable to enhance ionization efficiency
Use signal summing always and scheduled MRM for large panels
Be careful with quadrupole de-resolution (monitor ion ratios in many samples)
RUSSELL P. GRANT
LABORATORY CORPORATION OF AMERICA® HOLDINGS, BURLINGTON, NC, USA
33
HOW TO ACHIEVE LOWER QUANTIFICATION LIMITS: PART 3: FINESSE…tying the pieces together
Free T3 and T4 in 2003 …adsorption!
Challenges
• Sensitivity - 1pg/mL LLOQ from 200µL sample (<200 fg on column)
• Adsorptive losses (>99%) in aqueous solutions
• Over-dilution in equilibrium dialysis (disrupting free/bound)
• Internal standardization (and calibration) after dialysis
• T3 and rT3 are isomer
• Fragmentation of T4 produces T3 and rT3 – “in-source” isobar (resolution if assaying together)
34
Thyroid Metabolites – Signal gain and Ionization
• Good News:
• 3-4 common product ions from 1 precursor, yet rare product ion in biochemistry
• Negative ion mode (less noise/interferences)
• Uncommon mass range for Small Molecules
Triiodothyronine (T3) Reverse Triiodothyronine (rT3)
Thyroxine (T4)
50 100 150 200 250 300 350 400 450 500 550 600 650m/z, Da
Intensity, cps
[M-H]-
649.5
[I]-
127.0
50 100 150 200 250 300 350 400 450 500 550 600 650m/z, Da
[M-H]-
649.5
[I]-
127.0
Intensity, cps
Inten
sity, cps
[M-H]-
775.5
[I]-
127.0
50 100 150 200 250 300 350 400 450 500 550 600 650 700 750m/z, Da
0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8Time, min
400
800
1200
1600
Inten
sity, cps
T3
rT3
T4
0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8Time, min
8000.0
1.6e4
2.4e4
3.2e4
Inten
sity, cps T3
rT3
T4
Late elution (MeOH)
Post Colum Addition
35
0.00
10.00
20.00
30.00
40.00
50.00
60.00
0 4 8 12 16 20 24
Dialysis Time (hours)
Con
cen
trati
on
Measu
red
(p
g/m
l)
T4 Dialysis Losses absent following addition in dialysate on BOTH sides of membrane
Time-course of Free T4 with Equilibrium Dialysis at 37C
Free Thyroxine (T4) – 96-well Equilibrium Dialysis
Sample: 200µL sample and dialysate volume – mitigate dilutionary disruption
Sample: Dextran Blue (2 MDa) – Well malfunction (T4 99.9% Bound) Carry-over mitigation (do not inject)
Dialysate: Isotonic physiological - non disruptive to free fraction
Dialysate: Heat denatured protein added – non disruptive and mitigate adsorption
Post dialysis: Pre-wet & pre-fill IS in MeOH into aspiration tip – mitigate adsorption
Calibration: Dialysate matrix plus T1 (iodothyronine)* - mitigate adsorption
* Van Houcke et al.: Free thyroxine conventional reference procedure, Clin Chem Lab Med 2011;49(8):Ad1–Ad5 36
Free Thyroxine (T4) Assay Data
LLOQ 2 pg/mL (2.5pmol/L, 50fg on column)
0.2 0.4 0.6 Time, min 0
100
200
300
400
500
600
700
Inte
nsi
ty, c
ps
ULOQ 100 pg/mL (128.7pmol/L)
0.2 0.4 0.6 Time, min
8000
1.6e4
2.4e4
3.2e4
Inten
sity, cps
Free Thyroxine
2
-3.4
0.1
-0.4
-1.3
50
-7.0
-1.2
0.6
-2.5
100
3.0
-0.3
3.7
2.1
2
5.0
5.1
4.4
4.8
50
6.2
5.2
4.1
6.0
100
5.5
2.2
5.5
4.8
Accuracy (%) Precision (%)
Conc.(pg/mL)
Intra - 1
Intra - 2
Intra - 3
Inter
0
5
10
15
20
25
0 5 10 15 20 25
Lab
Co
r p
(p
g/m
L)
International RMP (pg/mL)
y = 1.0948x - 3.0391 r = 0.91
IA (pg/mL)
LC-M
S/M
S (
pg
/mL
)
5
10
15
20
25
30
35
5 10 15 20 25
y = 1.8145x – 4.407 r = 0.95
37
GnRH…decoupling selectivity and sensitivity
Challenges
• Sensitivity - 10pg/mL LLOQ from 50µL sample (<200fg on column)
• Adsorptive losses (>90%) in aqueous solutions
• First quantitative peptide assay for clinical diagnostics (2005)
• Unstable in specimens
• Isotopic separation of IS from analyte?
Good news
• Comparator immunoassay
• Its good to try to be first
• “labeled peptide” synthesis capability (malibu!)
38
39
Decoupling Selectivity and Sensitivity with 2D LC
79th Injection 1.2e5
Time, min
Inten
sity, cps
0.2 0.4 0.6 0.8 1.0 1.2 1.4
1st Injection 3.7e5
32nd Injection 2.4e5
4.0e4
8.0e4
1.2e5
Time, min
Inten
sity, cps
0.2 0.4 0.6 0.8 1.0 1.2 1.4
1.3e5
2.5e5
3.7e5
Time, min
Inten
sity, cps
0.2 0.4 0.6 0.8 1.0 1.2 1.4
6.0e4
1.2e5
1.8e5
2.4e5
pGlu
His
Trp
Ser
Tyr
Gly
Leu
Arg
Pro
Gly NH2
GnRH
Inten
sity, cps
0.2 0.6 1.0 1.4 1.8 2.2 2.6 3.0 3.4 3.8
Time, min
0
1000 2000 3000 4000 5000 6000 7000 8000 9000
GnRH Neat Calibrator
Ion Ratio ~1.2
Inten
sity, cps
0.2 0.6 1.0 1.4 1.8 2.2 2.6 3.0 3.4 3.8 Time, min
2.0e5 6.0e5
1.0e6 1.4e6
1.8e6 2.2e6
2.6e6
3.0e6 3.4e6
3.8e6 GnRH
ACN Plasma PPT
MeOH Plasma PPT
Water Injection
Post Column Infusion
Ref: Wagner A,Grant R , Poster, ASMS, 2007
40
RP Column 2 Time, min
Inten
sity, cps
0.2 0.4 0.6 0.8 1.0 1.2 1.4
RPLC
Max. 2.1e4 cps.
5000
1.3e4
2.1e4
HILIC Column 2 Time, min
Inten
sity, cps
0.2 0.4 0.6 0.8 1.0 1.2 1.4
HILIC
Max. 3.8e5 cps.
6.0e4
2.2e5
3.8e5
12x Increase
GnRH Elutes at <5% Organic in RPLC but 95% Organic in HILIC mode
Decoupling Selectivity and Sensitivity
2D RP - HILIC Time, min
Inten
sity, cps
0.2 0.4 0.6 0.8 1.0 1.2 1.4
5000
2.0e4
3.5e4 Plasma
Standard 100pg/mL
HILIC Only Time, min
Inten
sity, cps
0.2 0.4 0.6 0.8 1.0 1.2 1.4
2000
8000
1.4e4 Plasma Sample
100pg/mL
GnRH?
RP Only Time, min
Inten
sity, cps
0.2 0.4 0.6 0.8 1.0 1.2 1.4 0
1600
3200 Plasma Sample
100pg/mL
GnRH
Selectivity derived in 1st Dimension, Sensitivity generated in 2nd dimension
41
FIGURES OF MERIT: 2D-LC FOR GNRH
Ref: Wagner A,Grant R , ASMS, 2007
What is special about <1pg/mL (<4pmol/L)?
• Estrogen suppression “controlled” (Often < LLOQ)
• Pre-pubertal onset
• Free Estradiol
• Salivary Estradiol?
Challenges
• Sensitivity - 1pg/mL LLOQ every day
• Neutral molecule
• Very common mass range for Small Molecules
• Good News:
• Extensively studied and many tools to explore
• 14 years working on it (gen 3)
Estradiol…additional orthogonality
42
In 2003 on an API4000
+
NaHCO3/NaOHpH = 10.5
Regiospecific
3 min at 60CAfter LLE
OH
HO
S OO
Cl
NCH3H3C
OH
O
S OO
NCH3H3C
m/z 171.0Non-selective
Matrix Blank
0.2 0.4 0.6 0.8
Time, min
2000
6000
1.0e4
1.4e4
Inten
sity, cps D6 Estradiol
Estradiol
0.2 0.4 0.6 0.8
Time, min
4000
8000
1.2e4Inten
sity, cps
Estradiol
D6 Estradiol
10pg/mL LLOQ
SST
Serum
Gel (barrier)
Red blood cells0.2 0.4 0.6 0.8
Time, min
800
1600
2400
3200
Inten
sity, cps
Serum Serum Separator
Tube
43
44
Derivatization, Yield and Suppression - zero sum game?
Efficiency sensitivity) Loss = 50% (95%) Matrix effects ~85%
Estradiol (substrate) to Dansyl Derivative (product) with LC + polarity switch Determine disappearance/appearance using “disease and dirty samples”
Add neat and derivatized IS post extraction – Recovery yield + Matrix Effects
0%
20%
40%
60%
80%
100%
120%
0 5 10 15 20
Re
cove
ry /
Yie
ld
Time (min)
0%
20%
40%
60%
80%
100%
120%
0 0.5 1 1.5 2 2.5
Recove
ry
/ Y
ield
Derivative Concentration (mg/mL)
O-
OHCH3
O
OHCH3
S OO
NH+
CH3 CH3
60 80 100 120 140 160 180 200 220 240 260 280 300 320 340 360 380 400 420 440 460 m/z, Da0.0
1.0e8
2.0e8
3.0e8
Inten
sity, cp
s
506.2
156.1
171.1115.0
442.3
442.3 (-SO2)
171.1156.1
427.3 (-SO2)
427.3
0.10 0.20 0.30 0.40 0.50 0.600.0
2.0e5
4.0e5
6.0e5
8.0e5
1.0e6
Inten
sity, cp
s
E2
0.70 0.80 0.90 1.00 1.10 1.20 1.30 Time, min
DC-E2
Measurement of Low Level Endogenous Biomarkers for Use in Clinical Diagnostics, Dee et al, ASMS 2014, Poster
Increasing MSMS transmission efficiency API5000
Estradiol APCI 400C, Precursor Ion scan
50 100 150 200 250 300
5.0e4
1.0e5
1.5e5
2.0e5
Inten
sity, cps
+MS2 (273.40) CE (35): 30 MCA scans from Sample 1 (E2 Product Ion Scan 273) of E2 Product Ion Scan 273.wiff (Heated Nebulizer) Max. 2.2e5 cps.
50 60 70 80 90 100 110 120 130 140 150 160 170 180 190 200 210 220 230 240 250 260 270 280 290 300m/z, amu
1.0e4
2.0e4
3.0e4
4.0e4
5.0e4
6.0e4
7.0e4
8.0e4
9.0e4
1.0e5
1.1e5
1.2e5
1.3e5
1.4e5
1.5e5
1.6e5
1.7e5
1.8e5
1.9e5
2.0e5
2.1e5
2.2e5
Inte
ns
ity
, c
ps
94.9
107.1
81.2121.0
135.293.0
108.7133.1
119.4123.0
83.3105.1
149.1161.2 273.0
69.0
67.0
159.1147.0
162.996.956.9
143.2189.4 255.1
91.1 131.155.1 174.9
79.4 157.1137.0104.5
111.3 185.2217.171.2 203.4
151.3 171.2 199.161.8
84.9 129.060.3 245.0215.1187.1
156.2117.2 165.1 231.4257.9191.2 195.0
152.7141.598.7 101.0169.258.4 72.3
Max. 2.2e5 cps.
m/z, amu
Product Ions of m/z 273
200 250 300 350 400 m/z, amu
4.0e5
8.0e5
1.2e6
1.6e6
273
255
OH
HO
[M+H]+
[M+H-H20]+
Inten
sity, cps
Max. 1.7e6 cps.
Estradiol APCI 600C, Precursor Ion scan
Max. 2.4e6 cps.
200 250 300 350 400 m/z, amu
6.0e5
1.2e6
1.8e6
2.4e6
Inten
sity, cps
255 [M+H-H20]+
HO
> 2 orders of magnitude increase in MS/MS transmission efficiency
Product Ions of m/z 255
Max. 1.2e8 cps.
50 100 150 200 250 300 m/z, amu
4.0e7
8.0e7
1.2e8 In
tensity, cp
s
159
133
255
HO
HO
45
Heat induced stress…for m/z 255 - 159
200 250 300m/z, amu
4.0e5
1.2e6
2.0e6
2.8e6
3.6e6
Inten
sity, cps
271253
289
213
[MH]+
[MH-H2O]+
[MH-2H2O]+
Precursor Ion Scan of DHEA Product Ion Scan of dehydrated DHEA
1.0e6
3.0e6
5.0e6
7.0e6
100 200 300m/z, amu
Inten
sity, cps
253
197
157
[MH-2H2O]+
2C13 DHEA has a m/z 255 – 159 transition, circulates at 300 – 1500 x Estradiol
Estradiol : 24 interferences (m/z 255 159/133) Estrone : 16 interferences (m/z 273 159/133)
* Development and Application of 2D-LC-MS/MS in Clinical Diagnostics, Grant et al, ASMS 2007, Poster
9.00
10.00
11.00
12.00
13.00
14.00
15.00
0.00 2.00 4.00 6.00 8.00 10.00
Sep
ara
tio
n C
olu
mn
2 (
k')
Separation Column 1 (k')
“Heart-cut”
EstroneEstradiol
DHEA
Estriol
0 2 4 6 8 109
15
14
13
12
11
10
Retentive Capacity (k’) on a 75 x 2mm C18 “Peptide Column”
Ret
enti
ve C
apac
ity
(k’)
on
a P
FP
r2= 0.6*
46
Underivatized estrogens (LLE-LC-LC-MS/MS)
LLE (0.5mL), 2D-LC (5 min cycle time), Heat assisted APCI, 350 – 850fg on column
ULOQ (500 pg/mL)
Intensity, cps
0.2 0.4 0.6 0.8 Time, min0
2.0e4
4.0e4
6.0e4
8.0e4
1.0e5
1.2e5
Estrone
ULOQ (500 pg/dL)
0.2 0.4 0.6 0.8 Time, min0
1.0e5
2.0e5
3.0e5
4.0e5
5.0e5
Intensity, cps
Estradiol
LOQ (2.5 pg/mL)
0.2 0.4 0.6 0.8 Time, min0
200
400
600
800
1000
1200
1400
1600
Intensity, cps
Estrone
LOQ (1 pg/mL)
0.2 0.4 0.6 0.8 Time, min
Intensity, cps
Estradiol
0
400
800
1200
1600
2000
Calibration Curve
y = 0.0139x + 0.0152 r = 0.9995
0 40 80 120 160 200 240 280 320 360 400 440 480Analyte Conc. / IS Conc.
0.0
1.0
2.0
3.0
4.0
5.0
6.0
7.0
Analyte A
rea / IS A
rea
Calibration Curve
y = 0.0234x + 0.0069 r = 0.9997
0 40 80 120 160 200 240 280 320360 400 440 480Analyte Conc. / IS Conc.
0.0
1.0
2.0
3.0
4.0
5.0
6.0
7.0
8.0
9.0
10.0
11.0
12.0
Analyte A
rea / IS A
rea
Inter-assay comparison (with extraction/RIA)
Extract RIA Results (pg/mL)
LC
-LC
-MS
/MS
Res
ult
s (p
g/m
L)
0
40
80
120
160
0 20 40 60 80 100 120 140
y = 0.9006x + 3.8190 R2 = 0.8970
Extract RIA Results (pg/mL)
LC
-LC
-MS
/MS
Res
ult
s (p
g/m
L)
0
50
100
150
200
250
300
0 50 100 150 200 250
y = 1.1392x – 2.8302 R2 = 0.9776
Inter-assay comparison (with extraction RIA)
Quantitative LC-MS/MS Analysis of Endogenous Biomarkers for Clinical Diagnosis in Endocrinology, Morr et al, Poster ASMS 2006 47
Estrone
Conc.(pg/mL)
Intra - 1
Intra - 2
Intra - 3
Inter
2.5
12.4
-0.3
3.6
4.8
250
1.1
9.2
6.3
5.5
500
-1.4
4.0
5.1
2.6
0.1
2.9
7.3
6.3
7.4
250
5.5
2.8
3.8
5.1
500
3.3
2.4
1.6
3.7
Accuracy (%) Precision (%)
1.0
-7.1
-7.4
4.7
-3.3
2.5
1.1
1.1
5.6
2.7
250
-1.1
6.4
0.6
1.9
500
0.4
6.5
1.4
2.8
1.0
4.7
4.4
4.7
7.4
2.5
4.8
6.3
5.6
5.2
250
2.8
1.6
2.6
3.9
500
1.2
4.3
3.5
4.1
Accuracy (%) Precision (%) Estradiol
Validation Statistics (2006) and CDC Standardization Phase 1 (Aug 2014)
Vesper et al., Steroids 82 (2014) 7–13, 11 by IA, 6 by MS, Sample at 14.1 pg/ml – results from 9.4 – 64.8 pg/mL
LabCorp vs CDC (phase 1) Slope = 1.021
R= 0.9970 Bias = -0.469%
48
49
APCI +ve; –H20
0.0 1.0 2.0 3.0 4.0 5.0 Time, min
Inten
sity, cps
E2
ESI +ve; PSE2 Derivative
0.0 1.0 2.0 3.0 4.0 5.0 Time, min
Inten
sity, cps
E2
ESI –ve; Na quench
1.0 2.0 3.0 4.0 5.0 Time, min
Inten
sity, cps
E2
Since 2014..API5500..Estrogen suppression samples (~1pg on column)
Combinations of SLE plates TICE plates Phospholipid depletion plates 6 Derivatization schemes 2D-LC DMS (+ dopants)
50
Selectivity with 7 degrees of Orthogonality Dansylated Estradiol Red top serum vs Serum Separator Tube
SLE
1D-LC
1.0 2.0 3.0 4.0
5.0e5
1.0e6
1.5e6
2.0e6
Inten
sity, cps
1.0 2.0 3.0 4.0
6.0e5
1.2e6
1.8e6
2.4e6
Inten
sity, cps
Max. 2.2e6 cps.
Max. 2.6e6 cps.
CV 40%
Bias 500%
CV 15%
Bias 40%
Max. 1.6e6 cps.
4.0e5
8.0e5
1.2e6
1.6e6
Inten
sity, cps
1.0 2.0 3.0 4.0
Max. 1.5e6 cps.
5.0e5
1.0e6
1.5e6
Inten
sity, cps
1.0 2.0 3.0 4.0
SLE / +Ab Enrich 60% “Efficiency”
1D-LC
SLE
2D-LC
Max. 2.8e6 cps.
1.2e6
2.0e6
2.8e6
Inten
sity, cps
1.0 2.0 3.0 4.0
8.0e5
1.6e6
2.4e6
3.2e6
Inten
sity, cps
1.0 2.0 3.0 4.0
Max. 3.3e6 cps.
CV 20%
Bias 200%
CV 7%
Bias 5%
Max. 2.2e6 cps.
5.0e5
1.0e6
1.5e6
2.0e6
Inten
sity, cps
1.0 2.0 3.0 4.0
Max. 2.1e6 cps.
5.0e5
1.0e6
1.5e6
2.0e6
Inten
sity, cps
1.0 2.0 3.0 4.0
SLE / +Ab Enrich 60% “Efficiency”
2D-LC
51
Part 3 summary…what we learned
“On-column” amount is important , not LLOQ (yoct0gram/attoliter)
Adsorption must be controlled at each step of the assay
Anchor to reference methods if at all possible
Use 2D-LC to decouple selectivity and sensitivity variables
Establish selectivity and stoichiometry of reagents in worst case specimens
Improved LLOQ often requires “recovery losses” or
While counter-intuitive, more degrees of selectivity may be optimal
0.10 0.20 0.30 0.40 0.50 0.60 Time, min
2.0e4
4.0e4
6.0e4
Inten
sity, cps
Estradiol Blank
0.10 0.20 0.30 0.40 0.50 0.60 Time, min
2.0e4
4.0e4
6.0e4
Inten
sity, cps
LLOQ: 15fg on column
©2013 Laboratory Corporation of America All rights reserved. 52
Pat Holland Stacy Dee
Meghan Bradley Erin Fagan
Yvonne Wright Bill Curtin Kyle Cahill Mary Morr
Andrew Wagner Matthew Crawford
Walt Chandler, Ph.D. William Slade, Ph.D. Chris Shuford, Ph.D.
Thank You!