Pierre Picard1, Gregory Blachon1, Jean Lacoursiere1, Serge Auger1, Alex Birsan1, Michael Jarvis2, Adrian Taylor2

1Phytronix Technologies Inc., Quebec, Canada 2AB SCIEX, Ontario, Canada

OVERVIEW

INTRODUCTION METHOD

Challenges for the Analysis of 25-hydroxyvitamin D3 RESULTS:

CONCLUSIONS

MP #12

The Laser Diode Thermal Desorption™ (LDTD) ionization source has been coupled to a mass spectrometer equipped with the SelexION™ differential ion mobility cell, enabling a high throughput capacity for the analysis of 25-OH-vitamin D2 and D3 in biological matrix, with sample-to-sample analysis time of 9 seconds. The endogenous isobaric compound 7α-hydroxy-cholesten-3-one is known to be an interferent with similar MS/MS transition as 25-OH-D3. Thermal desorption process vaporizes all compounds simultaneously and isobaric molecules with similar structure may potentially interfere. Improvement of the analysis specificity is achieved by the action of the Differential Mobility Spectrometer. Desorption of individual compounds at the optimized DMS parameters shows specificity equivalent to liquid chromatography but at the speed of electronic separation, in milliseconds.

LDTD™ Ionization Source:

The LDTD uses a Laser Diode to produceand control heat on the sample support(Figure 1) which is a 96 wells plate. Theenergy is then transferred through the sampleholder to the dry sample which vaporizesprior to being carried by a gas in a coronadischarge region. High efficiency protonationwith strong resistance to ionic suppressioncharacterize this type of ionization, and arethe result of the absence of solvent andmobile phase. This allows for very highthroughput capabilities of 9 second sample-to-sample analysis time, without carry over.

SelexION™ technology:

The SelexION™ technology is a DifferentialMobility Spectrometer (DMS) placed in frontof the inlet of the mass spectrometer. Theionized molecules travel into the orthogonalgeometry shaped DMS(Figure 2).

� Use the dependence of ion mobility K on electric field strength E

� High-field mobility K(E) and low-field mobility K(0) are molecule dependant

� Molecular Trajectories are Electronically tunable

� Drift gas composition influences mobility K

� Any ion can be steered back onto the center-line, by application of a compound-specific DC Compensation Voltage(CV)

� Result: Increased Selectivity

� Use the dependence of ion mobility K on electric field strength E

� High-field mobility K(E) and low-field mobility K(0) are molecule dependant

� Molecular Trajectories are Electronically tunable

� Drift gas composition influences mobility K

� Any ion can be steered back onto the center-line, by application of a compound-specific DC Compensation Voltage(CV)

� Result: Increased Selectivity

α>0

α<α<α<α<0

αααα=0

t1

t2

)0()0()(

)(K

KEKE

−=αCompensation Voltage (COV)

Separation Voltage (SV)

Compensation Voltage (COV)

Separation Voltage (SV)

ions in transport gas

to mass spectrometer

� Common precursor ion masse (m/z 401)

� Similar fragmentation patterns

25-hydroxy vitamin D3, C27H44O2 7α-hydroxy-4-cholesten-3-one(7αC4), C27H44O2

Test and manipulations

• Calibration curve is prepared using a multilevel calibrator set from Chromsystem. Additional curve levels are prepared by dilution of calibrator with stripped serum.

• Matrix effect is evaluated by first measuring the original level of 6 different plasma samples and spiking them with a known amount of 25-OH-D3. We should observe constant endogenous + spiked quantity.

• Specificity is measured as the blank level peak area compared to the signal area at the limit of quantitation.

• Reproducibility is tested as intra-day and inter-day repeated measurements.

• Stability is evaluated in matrix, extract and dried onto Lazwell plate at room temperature and 4º Celsius.

• Finally, 6 concomitant drugs (Caffeine, Acetaminophen, etc.) are spiked in a QC to verify potential interferences

Effect of DMS on extracted matrix Effect of DMS on neat compound: 7αC4

Blank levels are less than 20% of the LOQ for both compounds. Figure 3a shows the effect of the DMS on the blank signal. Figure 3b presents the elimination of the 7αC4 interference tested by desorbing pure compounds and monitoring the 25-OH vitamin D3 trace. Quantitation curve ranges from 1-65 ng/mL and 1.5-94 ng/mL for 25-OH-vitamin D3 and D2 respectively. To assess the accuracy and precision, calibration points and QC’s are analyzed in triplicate. Reproducibility for n=3 is ranging from 0.6 to 12.3 %. Calculated concentrations of QC’s are within 15% of the reported values. Correlation between LC-MS/MS and LDTD-MS/MS samples is expressed by r2=0.952. Analyzing sample in 9 seconds with LDTD-MS/MS meets the guidelines of regulatory environment for all validation parameters.

Plasma analysis Results

Figure 1 Schematic of the LDTD ionization source.

Figure 2

Instrumentation• LDTD model S-960, Phytronix Technologies• QTRAP® 5500 Systems with SelexION™ technology, AB Sciex

MS Parameters• APCI (+) • Scan time : 0.020 s•DP: 80•CE : 15 eV•COV: 7.2 V•SV: 4200 V •DMO: -10 V•MRM :

�25-OH vitamin D3: 401.3 -> 365.2 �25-OH vitamin D3-d6: 407.3 -> 371.2�25-OH vitamin D2: 413.3 -> 337.3

LDTD Parameters• Laser power pattern :

� Increase laser power to 45 % in 3.0 s� Decrease laser power to 0 %

• Carrier gas flow : 3 L/min (Air)• Deposited sample volume: 2 µL

• LDTD ion source coupled to SelexION™ DMS achieves specific analysis of 25-OH vitamin D2 and D3 in the appropriate concentration ranges

• SelexION™ DMS increases selectivity and helps in removing isobaric interferences

• LC-MS/MS and LDTD-MS/MS measured values on real samples correlate within 95% confidence interval of statistical analysis

• LDTD provides the High-Throughput analysis of 25-OH vitamin D2 and D3 in 9 seconds sample-to-sample

expected mean sd cv accuracy

Std 1 1,5 1,8 0,2 12,8 120,8

Std 2 3,0 2,9 0,6 20,3 97,7

Std 3 5,9 5,0 0,6 11,0 84,4

Std 4 11,8 12,6 0,8 6,0 106,9

Std 5 23,6 19,7 1,8 9,0 83,7

Std 6 47,1 45,7 2,8 6,1 97,1

Std 7 94,2 98,1 6,8 6,9 104,1

Quality Control 38,2 41,9 109,7

Quality Control 16,5 16,2 98,1

Quality Control 66,0 63,2 95,7

Quality Control 42,7 49,4 115,6

25 Hydroxy vitamin D3

25 Hydroxy vitamin D2

expected mean sd cv accuracy

Std 1 1,0 0,9 0,2 17,5 86,0

Std 2 2,0 2,1 0,2 11,2 101,3

Std 3 4,1 4,4 0,6 13,9 107,4

Std 4 8,1 9,1 0,3 3,7 111,6

Std 5 16,3 14,9 0,0 0,2 92,0

Std 6 32,5 32,2 1,5 4,6 99,0

Std 7 65,0 65,1 2,4 3,7 100,1

Quality Control 22,4 20,9 93,3

Quality Control 92,5 86,2 93,1

Quality Control 63,1 60,4 95,8

Quality Control 27,6 27,2 98,5

Sample preparation

• Protein precipitation with 20 µl of plasma and 40 µl of MeOH containing internal standard

• Vortex 10 seconds

• Centrifuge 4 minutes at 14000 rpm

• Transfer 40 µl of supernatant

• Add 20 µl water NaCl saturated and 40 µl Hexane

• Vortex 30 seconds

• Wait 1 minute for phase separation

• Transfer 2 µl of the upper layer in Lazwell plate

• Analyze after complete solvent evaporation

Figure 3a Figure 3b

Additional LC-MS/MS vs LDTD-MS/MS correlation data

• Samples measured by LC-MS/MS using an established, and validated, method at the Toronto General Hospital

• The passing-Bablok regression revealed no significant deviation from linearity (Cusum test, P=0.10)

• Bland and Altman plot shows that the mean bias of the two methods was -0.885 and all samples are within the confidence interval of 95%

• LDTD-MS/MS analysis reproducibility is ranging from 1.3 to 13.8% (n=3)

LC-MS/MS LDTD-MS/MS Reproducibility Difference

(nM) (nM) (n=3) %

115 106,1 1,3 -7,7

140 125,7 3,7 -10,2

128 114,1 2,9 -10,9

80,1 71,7 13,8 -10,4

64,2 62,3 2,5 -2,9

42,6 48,9 1,4 14,7

65,1 70,8 7,8 8,8

72,5 68,3 7,3 -5,9

55,9 65,9 7,9 18,0

64,5 71,5 4,8 10,9

53,4 47,9 3,6 -10,3

54,8 60,0 4,0 9,5

29,1 33,5 9,8 15,3

128 123,9 6,2 -3,2

48,2 50,2 6,7 4,2

20,3 20,7 8,7 1,7

119 115,9 2,5 -2,6

94 90,5 2,3 -3,8

72 76,5 5,1 6,2

45,1 49,5 3,1 9,9