HOW TO ACHIEVE LOWER QUANTIFICATION LIMITS · 2017. 8. 18. · 1 . HOW TO ACHIEVE LOWER...

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


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


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


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


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


3X Pre-wet

Addition Step Addition Step Addition Step

IS Peak

Area


1X Pre-wet


Baseline


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


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%


Post-column “Orange versus Red peak” Analyte Area CV = 74%


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


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)

http://en.wikipedia.org/wiki/File:Estradiol.svg

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!

HOW TO ACHIEVE LOWER QUANTIFICATION LIMITS · 2017. 8. 18. · 1 . HOW TO ACHIEVE LOWER...

Documents

Transcript of HOW TO ACHIEVE LOWER QUANTIFICATION LIMITS · 2017. 8. 18. · 1 . HOW TO ACHIEVE LOWER...