Solubility%20Screening%20Measurements%20at%20Organon%20Newhouse - PowerPoint PPT Presentation

Title: Solubility%20Screening%20Measurements%20at%20Organon%20Newhouse


1
Solubility Screening Measurements at Organon
Newhouse

Darren Edwards Organon Laboratories, Newhouse
2
Outline

• Brief introduction to Organon Newhouse
• Requirements for a solubility screen
• How we screen for solubility SolKin
• Validation of SolKin
• Description of what was done
• Issues and problems
• Quality control
• Future work
• Summary

3
Organon

• Founded 1923 in the Netherlands
• Part of AKZO Nobel
• Organon has approx 13,000 employees, 2000 in RD
• Opened site at Newhouse in Scotland in 1947
• Over 300 scientists of all disciplines
• Responsible for Cardiovascular, CNS and Analgesia
  research at Organon

4
Physico Chemical Analysis at Newhouse
Organon Scotland
Lead Finding Teams
Proof of Concept Teams
Venture Teams
Lead
Pre-clinical / CMC
Pre-clinical / CMC
Marketing
Hit
Target
Registration
Optimization
Development
Sales
Development
Optimization
Clinical
Clinical
I
II
b
Development
Development
II
a
III
Exploratory
Full Development
Discovery
Development
and Launch
Research
Development
Physchem assays in support of HO and LO projects
5
Physicochemical tests at Newhouse

• HO and LO
• Solubility
• Screens (from DMSO)
• Solids (miniaturised shake-flask)
• Potentiometric (pSOL)
• Lipophilicity
• HPLC
• Miniaturised shake-flask
• pKa
• Sirius GLpKa
• PAMPA
• Via NV Organon, Oss

6
Physicochemical tests at Newhouse

• HO and LO
• Solubility
• Screens (from DMSO)
• Solids (miniaturised shake-flask)
• Potentiometric (pSOL)
• Lipophilicity
• HPLC
• Miniaturised shake-flask
• pKa
• Sirius GLpKa
• PAMPA
• Via NV Organon, Oss

Methodology Validation Quality control
7
Solubility screen - requirements

• Not high throughput
• Low 100s of compounds prepared per week
• Screen all for solubility
• Wanted to assay from 10 mM DMSO solution
• Automatically prepared for all compounds
• Maximum volume ca. 50 µl
• Chemists wanted a solubility value, rather than a
  range
• Sensitive (down to 1 mg/l)
• Reasonably accurate and precise
• Specific

8
SolKin assay

• Our solution the SolKin assay
• Based upon shake-flask technique
• HPLC analysis using conventional equipment
• Developed and validated in-house

9
Schematic of SolKin method
Sample Preparation (in duplicate)
Standardpreparation
Inject two different volumes(5 and 50 µl) and
quantify
All done in two 96-well platesOne for samples,
one for standards
10
Schematic HPLC instrumentation
Agilent 1100 system
WATER 0.1 Formic acid
VacuumDe-gasser
1 ml/min
Binary pump
Thermostatted autosampler
VacuumDe-gasser
MeCN 0.1 Formic acid
HPLC column, 5 cm Xterra C18, 3.5 µm
UV/vis PDA
Dionex Chromeleon and Excel
11
HPLC conditions
Gradient program
Typical Chromatogram
95
Acetonitrile
5
1
2
3
4
12
SolKin Method - Key Points

• Based upon a shake-flask approach
• Uses standard HPLC equipment and a generic
  reversed-phase gradient method with UV detection
• 230 nm usually used for quantitation
• 50 µl of 10 mM DMSO solution needed (lt0.5 mg)
• 12 µl for samples
• 25 µl for standard
• Specific (chromatographic)
• 2 DMSO in final solution
• Upper limit of 100 mg/l for MW 500.

13
Validation Work

• No guidelines for validating this sort of test
• Some very good examples in literature, but
  procedures used generally variable
• This is what we did

14
Validation Work

• Comparison of SolKin data with literature
• Recovery work
• Effects of DMSO on solubility
• Effects of agitation time

15
Validation Work

• Comparison of SolKin data with literature
• Recovery work
• Effects of DMSO on solubility
• Effects of agitation time

16
Comparison with Literature Data

• Compound Selection
• Literature solubility values variable
• Often no conditions quoted (purity, test media,
  ionic strength, temperature)
• Different amounts of DMSO used, different
  techniques gave different values
• Considered that literature solubility values for
  neutral compounds would be more reliable
• no influence of pH upon solubility
• Difficult to find poorly soluble (lt 100 mg/l?)
  neutral, available drug compounds with literature
  values
• most drugs higher than this

17
Compound Selection

• 15 marketed drugs (all except two effectively
  neutral)
• Structures and purity checked
• Only used compounds for which several literature
  sources agreed
• Values as measured from solid material no DMSO
• Triplicate measurements (3 different occasions,3
  analysts) in phosphate buffered saline (PBS)
• pH 7.4
• 0.05 M (no consensus in literature as to which
  buffer strength to use 0.001 M to 0.1M!)
• 0.15M NaCl used in buffer (again no consensus in
  literature, but most common)
• Ambient temperature (measured as 211C)

18
Selected Compounds
haloperidol
progesterone
griseofulvin
estradiol
phenytoin
testosterone
Hydrocortisone-21-acetate
spironolactone
digoxin
triamcinolone
diazepam
thiamylal
dexamethasone
lorazepam
prednisolone
19
SolKin data versus Literature

• Is this comparison useful?
• How good is literature data?
• Highlights lack of published data

20
Validation Work

• Comparison of SolKin data with literature
• Recovery work
• Effects of DMSO on solubility
• Effects of agitation time

21
Validation - Recovery

• Question Is a three point calibration curve
  produced by an autosampler sufficient to give the
  required accuracy?
• Our solution analyse aqueous solutions of three
  water soluble drug compounds prepared at a range
  of different concentrations
• Check concentration found against prepared
  concentration (in duplicate)
• Also, check concentrations after filtering
• Compounds used

• Methyl parabens

• Salicylic acid

• Caffeine

22
Validation - Recovery Results
23
Recoveries Summary

• Limited experiment..
• Generally recoveries ca. 90
• Some reduction in recoveries at high
  concentrations (175 mg/l)
• Probably due to large extrapolations from
  calibrated range
• Upper limit 100 mg/l (MW 500)
• Methyl parabens shows evidence of membrane
  retention
• Filter plates chosen after discussions with
  manufacturer (Millipore)
• Interested to hear what plates other groups
  use..
• Lipophilic compounds more of a problem?
• Recoveries considered acceptable for solubility
  screen

24
Validation Work

• Comparison of SolKin data with literature
• Recovery work
• Effects of DMSO on solubility
• Effects of agitation time

25
DMSO/solubility

• Initially developed test with 2 DMSO
• Common value in literature
• Wanted to see how much difference 5 DMSO would
  make
• 11 compounds in duplicate
• 2 DMSO upper limit for MW 500 100 mg/l
• 5 DMSO upper limit for MW 500 250 mg/l

26
DMSO/solubility
2 DMSO chosen100 mg/l upper limit considered
acceptableLimited dataset!
27
Validation Work

• Comparison of SolKin data with literature
• Recovery work
• Effects of DMSO on solubility
• Effects of agitation time

28
Agitation Time

• Several differences in literature as to stirring
  time used
• Anything between 1 and 48 hours
• Investigation made using agitation times of 1.5
  and 24 hours
• 1.5 commonly used
• 24 longest practical stirring time for solubility
  screen
• Duplicate samples, results averaged.

29
Agitation Time
Results after 1.5 hours consistently higher than
those after 24 hours agitation. 24 hours gives
better correlation to literature.Chose 24 hours,
but 1.5 hours OK for screen?
30
Quality Control
31
Quality Control

• Samples run in duplicate.
• 48 per plate
• 2 standards (estradiol and haloperidol) run in
  duplicate for every 22 (maximum) samples.
• Estradiol near lower limit of assay (solubility
  ca. 1.5 mg/l)
• Haloperidol, pKa 7.7, highlights effects of pH
  changes
• Used to generate process control charts..
• Highlight the presence of special causes
  which show if the process is out of control

32
Quality Control Process Control Diagram

• Using data to try and generate acceptance
  criteria for assay and give idea of variability
• Variation can be detected and data rejected if
  necessary.
• Work ongoing

Minitab software
Special cause 9 points below mean
33
Issues identified from validation and QC work

• Still learning and validation ongoing, several
  key issues..
• Lack of high quality literature data
• Often no indication of conditions
• Could we set up a database of solubility data (
  conditions?) on marketed drugs via this forum?
• No consensus on minimum validation for this sort
  of test
• Is it possible to use a forum such as this to
  come up with guidelines?
• Quality control
• No guidelines
• What is acceptable variation?
• Forum guidelines for this?

34
Future work

• Investigating the use of very fast and
  multi-channel LCs to decrease sample analysis
  times
• Refining quality control procedures
• Use robotic workstation for sample preparation
• Manually with multi-channel pipettes at present
• Mass Spectrometry as detection
• Other pH values
• e.g. pH 6.5 to allow MAD analysis to be done
  (with PAMPA data from Organon NV, Oss)
• Investigate other media
• e.g. Pharmacology in-vitro test solutions

35
Summary

• Described our solubility screening assay
• Modified shake-flask method
• Uses 10 mM dmso solutions
• Analysis uses standard HPLC equipment
• Medium throughput
• Described our validation/quality control work
• Highlighted several issues that we found
  developing/validating a solubility method

36
Acknowledgements

• Organon Labs, Newhouse
• Wullie Arbuckle
• Yvonne Lamont
• Strathclyde University
• George Gettinby, Professor of Statistics