Ronald de Vries | EBF Meeting Barcelona | 19-21 November 2014

Combined use of LBA + LC-MS/MS in drug development of a

2 kDa peptide: 1+1=3 or where complementary data made a difference

Pictured above: The structure of HIV.

LBA versus LC-MS/MS assay

Basic principles of the analytical techniques differ significantly

•  LC-MS/MS tends to measure parent peptide drug only

−  active and inactive metabolites are missed unless intentionally added to the assay

•  LBA measures

–  peptide drug only OR

–  peptide drug + active metabolites OR

–  peptide drug + active metabolites + inactive metabolites

Sandwich ELISA

Most common Ligand Binding Assay conformation

Sandwich ELISA measuring drug, active metabolite and inactive metabolite

YY

YY

Peptide drug - detected

Inactive metabolite - detected

Active metabolite - detected

YY

Sandwich ELISA measuring drug but no metabolites

YY

YY

Peptide drug - detected

Inactive metabolite – NOT detected

Active metabolite – NOT detected

YY

The draft FDA Bioanalytical Method Validation Guidance (2013)

“LBA assays should be compared with a validated reference method (such as LC-MS) using incurred samples and predetermined criteria should be used to assess the accuracy of the LBA method”.

•  This text suggests that an LBA and an LC-MS/MS assay should have a comparable result in order to be valid.

•  However, based on the difference in basic principles of both assay formats, the results from both assays can differ significantly …

•  Difference in results LBA vs LC-MS/MS should NOT be an issue, but it is important to understand “what” each assay is measuring

An example from Janssen Portfolio

Different results from PK assays

•  LBA (sandwich ELISA) used as bioanalytical method before in-licensing of this peptide drug

• When using LC-MS after in-licensing

−  concentrations in dog, rat and human plasma/serum by LCMS were much lower than by LBA

−  difference was larger at later time points

−  cross-reactivity LBA with metabolites suspected

•  Samples from a Dog Study after single IV dosing, same samples analyzed by LCMS and ELISA

SubjectSample

ID Hour MinuteLCMS

(ng/ml)ECLIA (ng/ml) ECLIA/LCMS

062M 324 2 42000 74470 1.8062M 325 5 10700 28042 2.6062M 326 15 647 4222 6.5062M 327 30 87.0 1191 13.7062M 328 1 55.1 384 7.0062M 329 1.5 12.7 137 10.8062M 330 2 7.12 17.3 2.4062M 331 4 BQL BQL162F 375 2 78700 138000 1.8162F 376 5 15500 47950 3.1162F 377 15 528 5171 9.8162F 378 30 65.8 1142 17.4162F 379 1 13.7 245 17.9162F 380 1.5 12.1 60.4 5.0162F 381 2 4.81 BQL162F 382 4 BQL BQL

Metabolite identification

•  Process rat, dog and human plasma/serum samples using

−  Protein precipitation with acetonitrile

−  Incubation with capture antibody used in ELISA, followed by isolation of IgG’s (including capture antibody) by protein G

•  Analyze by LC-MS (Q-TOF, Synapt G2-S)

•  Clean up spectra by ion extraction based on charge state (next slide)

Ion extraction based on charge state

original chromatogram

10 Bioanalysis (2012) 4(5), 595 – 604

Ion extraction based on charge state

original chromatogram

3+  2+  

Bioanalysis (2012) 4(5), 595 – 604

Ion extraction based on charge state

original chromatogram 3+ charge state filtered chromatogram

12

3+  2+  

0.334  Da    

Bioanalysis (2012) 4(5), 595 – 604

Ion extraction based on charge state

original chromatogram 3+ charge state filtered chromatogram

13

3+  2+  

0.334  Da    

Bioanalysis (2012) 4(5), 595 – 604

D

R

R

A

E

P D N Y C

D

F V

Q

S

C

K

A

D

R

R

A

E

P D N Y C

D

F V Q

S

C

K

A

D

R

R

A

E

P D N Y C

D

F

V Q

S

C

K

A

D

R

R

A

E

P D N Y C

D

V Q

S

C

K

A

R

R

A

P D N Y C

D

S

C

K

R

R

P D N Y C

D

S

C

K

D

R

R

A

E

P D N Y C

D

F

S

C

K

D

R

R

A

E

P D N Y C

D

S

C

K

R

R

A

P D N Y C

D

F

S

C

K

R

R

P D N Y C

D

F

S

C

K

R

P D N Y C

D

F

S

C

K

R

P D N Y C

D

S

C

K

Metabolism scheme – 11 metabolites identified

Protein precipitation vs Protein G

Rat IX8 day6 10 min

Time2.00 4.00 6.00 8.00 10.00 12.00 14.00 16.00 18.00 20.00

%

0

100

2.00 4.00 6.00 8.00 10.00 12.00 14.00 16.00 18.00 20.00

%

0

100

COR1-025AFAMMA-3POS TOF MS ES+ TIC

3.27e721.14

5.652.49

5.204.56

13.6413.019.09

18.76

COR1-006AFAMMA-3POS TOF MS ES+ TIC

2.56e621.00

5.53

4.4812.85

8.96

13.52

18.04 18.64

PP

Protein G

Chromatograms obtained after ion extraction of +3 charge state

Identity of metabolites versus retention time

Rat IX8 day6 10 min

Time2.00 4.00 6.00 8.00 10.00 12.00 14.00 16.00 18.00 20.00

%

0

100

2.00 4.00 6.00 8.00 10.00 12.00 14.00 16.00 18.00 20.00

%

0

100

COR1-025AFAMMA-3POS TOF MS ES+ TIC

3.27e721.14

5.652.49

5.204.56

13.6413.019.09

18.76

COR1-006AFAMMA-3POS TOF MS ES+ TIC

2.56e621.00

5.53

4.4812.85

8.96

13.52

18.04 18.64

PP

Protein G

R R A

P D N Y C

D S C K

R R

P D N Y C

D S C K

D

R R A E

P D N Y C

D

V Q

S C K

A

LBA on HPLC fractions

Rat IX8 day6 10 min

Time2.00 4.00 6.00 8.00 10.00 12.00 14.00 16.00 18.00 20.00

%

0

100

2.00 4.00 6.00 8.00 10.00 12.00 14.00 16.00 18.00 20.00

%

0

100

COR1-025AFAMMA-3POS TOF MS ES+ TIC

3.27e721.14

5.652.49

5.204.56

13.6413.019.09

18.76

COR1-006AFAMMA-3POS TOF MS ES+ TIC

2.56e621.00

5.53

4.4812.85

8.96

13.52

18.04 18.64

PP

LBA

Conclusions - Project

•  For the peptide drug, a large difference between LBA and LC-MS/MS was observed

•  It was shown that the difference was due to cross reactivity in the LBA with metabolites of the peptide drug

•  Tools used to demonstrate cross-reactivity with metabolites in general and showing with which metabolites the cross-reactivity occurred were

−  Incubation with capture Ab followed by Protein G purification

−  LBA on HPLC fractions

• Metabolites of the peptide drug were identified, and ion extraction based on charge state was a very useful tool to clean up the spectra to aid in metabolite ID of the peptide

Conclusions - Strategic

•  Bioanalysis using LBA and LC-MS/MS assays can have a significantly different result because of cross reactivity in LBA and/or because of LC-MS/MS only measuring the peptide drug and not active metabolites (unless added intentionally)

•  The use of both platforms is very useful to gain a better understanding of the peptide drug and of the read-out of the assays – could be part of a company strategy to use both assays for this purpose

•  In contrast with what the DRAFT FDA Guidance is suggesting, a difference in results LBA vs LC-MS/MS should NOT be an issue, but it is important to understand “what” each assay is measuring

Acknowledgements

• Damien Fink

•  Tony Greway

•  Luc Sips

•  Philip Timmerman

• Marc Verhemeldonck

•  Filip Cuyckens