Measuring pKas, logP and Solubility by Automated titration

Sirius Analytical Measuring pK a s, logP and Solubility by Automated titration Jon Mole Technical Sales Manager www.sirius-analytical.com

Contents Introduction to Sirius Overview of instrumentation & assays for pK a , logP/D Validation studies Principle of our “CheqSol” Solubility method Our early theories Four classes of solubility behaviour - implications Modelling gastrointestinal precipitation/dissolution Conclusions / 70

An introduction to Sirius Sirius was founded in 1990 in the UK. We are a manufacturer and vendor of instrumentation for measurement of physicochemical parameters. We also run an Analytical Service , and measure thousands of samples for hundreds of customers, worldwide, each year. / 70

Sirius locations Sirius Analytical Ltd. Company headquarters in UK Manufacturing Engineering Software Chemistry R & D Administration Located in Forest Row, East Sussex 30 minutes from London Gatwick Airport Direct sales in some countries, distributors in others Sirius Analytical Inc. Support for North American customers Instrument service Installation Training Sales Stock of parts Located in Lakewood, NJ 60 minutes from Newark Airport / 70

What we do We make instruments for measuring physicochemical properties of ionizable compounds pK a logP/D Solubility Dissolution Widely accepted assays regarded as “gold standard” Instruments installed at most major pharmaceutical companies We also offer an analytical service for these parameters (and many more) / 70

pK a is the pH at which an ionisable group is “half-ionised” / 70 Propranolol (a base): pK a = 9.53 BH + B BH + B Flumequine (an acid): pK a = 6.27 HA A - HA A -

Most drugs ionize in solution Other properties (lipophilicity, solubility, permeability) are pK a -dependent In general: neutral molecules are more easily absorbed by membranes ionized molecules remain in plasma and are predominantly cleared by renal excretion Other reasons: Useful in formulation and salt selection. W ith knowledge of pK a values, the species which is stable over the largest pH range can be selected. Why is pK a important? / 70

/ 70 2 or more acidic groups, no basic ~ 3% 1 basic group, no acidic ~ 42% 1 acidic group, no basic ~ 12% Others ~ 3% 1 basic group + 2 or more acidic ~ 3% 1 acidic group + 2 or more basic ~ 4% 1 acidic group + 1 basic ~ 8% 2 or more basic groups, no acidic ~ 25% With thanks to Tim Mitchell and Ryszard Koblecki, Millennium Pharmaceuticals Ltd. 32,437 Ionizable drugs in World Drug Index (63% of total)

Human Gastrointestinal (GI) Tract / 70 STOMACH 0.1 m 2 DUODENUM 0.1 m 2 JEJUNUM 60 m 2 ILEUM 60 m 2 COLON 0.3 m 2 pH (fasted) 4.6 (2.4 - 6.8) 6.1 (5.8 - 6.2) 1.7 (1.4 -2.1) 6.5 (6.0 - 7.0) 6.5 8.0 5.0 - 8.0 pH (fed) 5.0 (0.1 hr) 4.5 - 5.5 (1 hr) 4.7 (2 hr) 6.5 8.0 3-4 h small Intestine transit time

LogP and logD describe lipophilicity / 70 P = partition coefficient. The ratio of concentrations of unionised species dissolved in two immiscible solvents (e.g. water + octanol) which are in equilibrium. D = Distribution Coefficient. The ratio of all species dissolved in two immiscible solvents which are in equilibrium. P is constant D is pH-dependent

Lipophilicity profiles are pK a and pH dependent / 70 Desipramine pK a = 10.14 Diphenhydramine pK a = 8.26 Triamterene pK a = 3.92 These molecules all have similar value for log D at pH 7.4. Their lipophilicity profiles are quite different Flat part of curve: log D = log P of neutral species Diclofenac pK a = 3.99 Phenobarbital pK a = 7.43 Nifuroxime pK a = 10.56 Big changes in lipophilicity occur over physiological pH range Physiological pH range

/ 70 Why is logP (and logD) important? LogP and logD (lipophilicity) provide a rough guide to pharmacokinetic behavior. LogD at pH 7.4 Implications for drug development Below 0 Intestinal and CNS permeability problems. Susceptible to renal clearance. 0 to 1 May show a good balance between permeability and solubility. At lower values, CNS permeability may suffer 1 to 3 Optimum range for CNS and non-CNS orally active drugs. Low metabolic liabilities, generally good CNS penetration 3 to 5 Solubility tends to become lower. Metabolic liabilities increase Above 5 Low solubility and poor oral bioavailability. Erratic absorption. High metabolic liability, although potency may still be high.

Why is solubility important? / 70 Poorly soluble molecules rarely make successful drugs They are difficult to absorb, to formulate and to analyze Discovery and lead optimization it helps in identification of potential screening and bioavailability issues it is valuable in planning chemistry changes Biopharmaceutical evaluation it is important for the confirmation of bioavailability issues during early trials of drugs, it is used in the design of animal formulations, as well as for human formulation design Development solubility knowledge is needed for biopharmaceutical classification, biowaivers and bioequivalence it is also required for formulation optimization and salt selection Manufacturing solubility affects the optimization of manufacturing processes

Solubility is also pK a and pH dependent / 70 Compounds are more soluble when ionized Lower plateau is the “intrinsic solubility” Propranolol Base pK a = 9.54 S 0 = 81 μ g/mL (314 μ M) Diclofenac Acid pK a = 3.99 S 0 = 0.9 μ g/mL (4.1 μ M)

SiriusT3 Our “next generation” system pK a logP/D Solubility Built in UV/Vis Sub-mg sample requirement Autoloader for 192 samples More automation, easier to use www.sirius-analytical.com/products/SiriusT3.shtml / 70

SiriusT3 - Dispenser Module 5 Mini dispensers Cosolvent 6 way valve Reagents storage Argon sparging UV/Vis diode array & detector / 70

SiriusT3 – Titrator Module Arm moves probes into calibration, wash and assay positions Built in turbidity sensor Peltier temperature control Argon flow Flowing water wash for optimum cleaning / 70

SiriusT3 – Autoloader Module 4 x 48 vial racks Standard footprint Robotic gripper arm to automatically move samples Built in ultra-sonic bath / 70

Titration cell for SiriusT3 Capillaries, for adding reagents pH electrode, diameter 3mm Glass vial, 4 ml total capacity Electronic thermometer Automatic overhead Stirrer Probes require a minimum of 0.5mL of solution contained in glass vial. Typical assay volume = 1ml. UV Dip Probe. / 70

pH-metric pK a – experimental process Set up assay SiriusT3 software includes templates for all assays Data Collection weigh sample into vial (or dispense stock solution) instrument adds water (or water-CoSolvent) instrument adjusts pH, then titrates with acid or base Calculation of results ( in software) pK a result obtained by analyzing the shape of the titration curve Calculation fully automated in SiriusT3 software Titration of flumequine in 39.8% methanol / 70

Example: vancomycin Measured pK a s: Van 4- + H+ VanH 3- pK a6 11.88 ± 0.01 VanH 3- + H+ VanH 2 2- pK a5 10.15 ± 0.01 VanH 2 2- + H+ VanH 3 - pK a4 9.28 ± 0.01 VanH 3 - + H+ VanH 4 pK a3 8.62 ± 0.01 VanH 4 + H+ VanH 5 - pK a2 7.48 ± 0.01 VanH 5 + + H+ VanH 6 2+ pK a1 2.64 ± 0.01 Difference Curve can handle 1, 2 or several pK a s ½Á (bound H¤ per Van) / 70 0 1 2 3 4 5 6 7 1 2 3 4 5 6 7 8 9 10 11 12 13 pH (concentration scale) Difference Curve 0 50 100 % Species 1 2 3 4 5 6 7 8 9 10 11 12 13 pH Distribution of species VanH 5

* D-PAS = Dip-Probe Absorption Spectroscopy pK a measurement by UV Multi-wavelength UV technique 220 to 750nm, diode array Fibre optic dip probe allows spectral measurement during titration Less sample required (down to 10 -6 M) 3uL of 10mM Stock equally sensitive over entire pH range pK a s measured below 1, above 13 Allows fast pKa measurement (just 4 mins) / 70

Set up assay SiriusT3 software has built-in templates for UV pK a assays Data Collection Prepare 10mM stock solution of sample in DMSO Pipette 3 μ L of stock solution into vial instrument adds water (or water + CoSolvent) & buffer instrument adjusts pH, then titrates with acid or base Calculation of results ( in software) Target Factor Analysis (TFA) method finds pK a s, even when they are overlapping or spectral change as a function of ionization is small Calculation fully automated in SiriusT3 software UV pK a - experimental / 70

3-D spectrum: pH vs. absorbance vs. wavelength pK a from UV spectra / 70

pK a result calculated by Target Factor Analysis (TFA) / 70 pKa Result

Dissolve sample in water-miscible CoSolvent + water Water-CoSolvent should be ionic-strength-adjusted to 0.15M with KCl Solvents supported fully: methanol (80%) 1,4 dioxane (60%) DMSO (60%) ethanol (60%) ethylene glycol (60%) DMF (60%) THF (60%) Acetonitrile (50%) MDM-mix (20% Methanol, 20% dioxane, 20% acetonitrile) (60%) Iso-propyl alcohol Using CoSolvents / 70

Using Sirius: Quinine hydrochloride pK a Sample weighed into vial SiriusT3 adds CoSolvent + water (45 wt% methanol), lowers pH and titrates with KOH ( green curve) SiriusT3 adds more water (now 35 wt% methanol), lowers pH and titrates with KOH ( red curve) SiriusT3 adds more water (now 25 wt% methanol), lowers pH and titrates with KOH ( blue curve) Results from Yasuda-Shedlovsky extrapolation: pK a 1 = 8.49 ± 0.02 pK a 2 = 4.22 ± 0.01 Example of a CoSolvent titration / 70

3-aminobenzoic acid : an example of an ordinary ampholyte Assigning pK a s to ionizable groups ...using CoSolvent p s K a titrations Yasuda-Shedlovsky slope direction: up (red) = acidic group down (blue) = basic group / 70

Detecting precipitation SiriusT3 has built in turbidity detection . Shaded area shows pH where sample precipitated. This is a warning, do not use this data to determine pK a of miconazole! Repeat in cosolvent to avoid precipitation and get reliable pK a data. / 70

Set up assay SiriusT3 software has templates for logP assays Data Collection weigh sample into vial (or use stock solution). instrument adds water and octanol instrument adjusts pH, then titrates with acid or base Calculation of results ( in software) logP result obtained by analyzing the shape of the titration curve (procedure requires pK a value) Calculation fully automated in SiriusT3 software pH-metric logP - experimental / 70

Principles of pH-metric logP measurement A solution of the sample is titrated in a two-phase system (water + octanol) The sample can ionise in water (pK a ), or it can partition into octanol (logP) The presence of the octanol disturbs the pK a equilibrium. The pK a shifts to a new value (p o K a ) to minimise this disturbance. We calculate the logP from this shift in pK a . / 70

Flumequine (acid) pK a = 6.27, p o K a = 7.99 log P = 1.72 Titrations with equal volumes of water and octanol / 70 Lipophilicity profiles: these profiles are correct for high logD, but do not show partitioning of ionic species Diacetylmorphine (base) pK a = 7.95, p o K a = 6.37 logP = 1.58 Aqueous pK a p o K a

Shake flask vs. pH-metric line calculated using log D equation for monoprotic base, using pK a = 9.54, log P 0 = 1.83, log P 1 = -1.32 (pH-metric data, 0.15M KCl, 25°C [8]) [8] Caron, G., Steyaert, G., Pagliara, A., Reymond, F., Crivori, P., Gaillard, P., Carrupt, P.A., Avdeef, A., Comer, J., Box, K.J., Girault, H.H., Testa, B. Helv Chim Acta . 82, 1211-1222 (1999) [9] Barbato, F., Caliendo, G., Larotonda, M.I., Morrica, P., Silipo, C., Vittoria, A. Farmaco . 45, 647-663 (1990) Pindolol points from shake-flask experiments, various buffers, 0.1M [9] / 70

The world’s most powerful system for measuring pK a SiriusT3 is our third generation instrument. We have spent over 15 years of continuous research and development on improving our assays and calculations for pK a , logP and solubility measurement. If a sample has a pK a between 2 and 12, we can always measure it on the SiriusT3 system. Every year we measure pK a of hundreds of samples that customers send us, and we never fail (provided the sample has a pK a and is chemically stable and pure). / 70

Validation of UV pK a method R.I. Allen, K.J. Box, J.E.A. Comer, C. Peake, K.Y. Tam, J. Pharm. Biomed. Anal., 17 , 699-712, 1998. K.Y. Tam, K. Takács-Novák, Pharm. Research,. 1999, 16, 374-381 R.C. Mitchell, C.J. Salter, K.Y. Tam, J. Pharm. Biomed. Anal., 1999, 20, 289-295 K.Y. Tam, M. Hadley, W. Patterson, Talanta, 1999, 49, 539-546 pK a (spec) = 1.006 x pK a (pH-metric) n = 31 R 2 = 0.999 RMSD = 0.098 Benzoic acid (3.98) Icotidine (3.29, 5.39, 6.22, 9.97) Lupitidine (2.79, 5.96, 8.25, 9.66) Nicotinic acid (2.10, 4.63) Nitrazepam (2.90, 10.39) Niflumic acid (2.28, 4.86) m-aminobenzoic acid (3.17, 4.54) p-aminosalicylic acid (1.79, 3.58) Phthalic acid (2.70, 4.86) Phenol (9.73) Phenolphthalein (8.87, 9.35) Pyridoxine (4.90, 8.91) Quinine (4.33, 8.59) SB-221789 (2.74) SKF-75250 (1.48, 6.59) (measured at 25ºC and an ionic strength of 0.15 M) / 70

Validation of pH-metric logP method B. Slater, A. McCormack, A. Avdeef and J.E.A. Comer, J. Pharm. Sci . 1994, 83 ,1280-1283 J.E.A. Comer, K. Chamberlain and A. Evans in J. Devillers (Ed.), SAR QSAR Environ. Res., Vol.3 Issue 4; Molecular Descriptors , Gordon and Breach, Philadelphia 1995, pp. 307-313. K. Takács-Novák and A. Avdeef, J. Pharm. Biomed. Anal. 1996, 14 ,1405-141; 61 samples over eight logP units amino acids, peptides, ampholytes, barbiturates, ß-blockers, herbicides, phenols, various others Graph plotted using Polyfit program from RefinementPro 2 / 70

CheqSol Technology – for Solubility measurement In 2004, Sirius introduced a new patented technology for measuring solubility of ionizable drugs. Recent investigations at Sirius using our CheqSol solubility assay has given us some interesting insights into supersaturation effects and the relationships between dissolution and precipitation rates for a range of drugs. Our latest research indicates that molecules can be placed into one of four classes: Chasers , Non-Chasers , Super-Dissolvers and Ghosts . / 70

Introduction to CheqSol Unique method for solubility measurement Runs on Sirius GLpKa, PCA200 & SiriusT3 instruments Requires pK a value Uses “Chasing Equilibrium” process to determine intrinsic solubility / 70

Sirius definitions of solubility Kinetic Solubility is the concentration of a compound in solution at the time when an induced precipitate first appears Equilibrium Solubility* is the concentration of compound in a saturated solution when excess solid is present, and solution and solid are at equilibrium Intrinsic Solubility ** is t he equilibrium solubility of the free acid or base form of an ionizable compound at a pH where it is fully un-ionized * also called Thermodynamic Solubility ** Hörter, D.; Dressman, J. B. Adv. Drug Deliv. Rev. , 1997, 25, 3-14 / 70

Starting the CheqSol Assay - Seeking precipitation Precipitation causes light scattering, and system detects this as an increase in the light absorbed. Kinetic solubility determined at point of precipitation. Before precipitation, no light is absorbed by the solution Absorbance 0.0 0.8 1.6 2.4 3.2 200 300 400 500 600 700 / 70 Solid added to vial. (5 to 20mg on GLpKa) (0.5 to 2mg on SiriusT3) Instrument adds water (or water-cosolvent), then adjusts pH to dissolve sample. Solution titrated towards the pH where the sample becomes neutral. Eventually it precipitates After precipitation, most light is scattered Wavelength (nm) 0.0 0.8 1.6 2.4 3.2 200 300 400 500 600 700

The Bjerrum Graph: a graphical view of solubility Bjerrum Graphs provide a graphical view of solubility They are theoretical curves plotted using pH, pK a , solubility and concentration of sample They are related to the Distribution of Species graph Graph below shows distribution of species of Pindolol in aqueous solution The next slide shows the Bjerrum function B j vs. pH / 70 B BH + pH 2 4 6 8 10 12 % Species 0 50 100 pK a = 9.54

Understanding the Bjerrum Graph / 70 Sample = base with one pK a Precipitate is present pH 0.0 0.5 1.0 2 4 6 8 10 12 pH = pK a Sample is unionised at this pH Sample is ionised at this pH Bj =Moles of bound H + ions per mole of sample Precipitation Bjerrum Graph. For a base with one pK a , Solution Bjerrum Graph. For a base with one pK a , 1.0 B BH + pH 2 4 6 8 10 12 % Species 0 50 100

/ 70 Sample = base with one pK a Understanding the Bjerrum Graph This distance depends on: Solubility (for a given concentration, distance increases as solubility decreases) Concentration (for a given solubility, distance increases as concentration increases) Precipitate is present pH 0.0 0.5 1.0 2 4 6 8 10 12 pH = pK a Sample is unionised at this pH Sample is ionised at this pH Bj =Moles of bound H + ions per mole of sample Precipitation Bjerrum Graph. For a base with one pK a , Solution Bjerrum Graph. For a base with one pK a , 1.0 pH 0.0 1.0 2 4 6 8 10 0.5 Sample = acid with one pK a Sample is ionised at this pH Sample is unionised at this pH

CheqSol example – solubility of Pindolol (a chaser) Pindolol is a beta-blocker, used to reduce hypertension It’s a secondary amine with pK a of 9.54 (25°C, 0.15M ionic strength) The neutral form B is poorly soluble in water at high pH The ionized form BH + is soluble at low pH BH + soluble B insoluble / 70

Pindolol - Chasing method For Pindolol, the kinetic solubility falls on a different Precipitation Bjerrum Graph to the rest of the data Intrinsic solubility is determined from the data points on the Equilibrium Precipitation Graph, as explained in the following slides Kinetic solubility is higher than Intrinsic solubility Assay took 37 minutes to measure kinetic and Intrinsic solubility / 70 pH 0.0 0.5 1.0 0 2 4 6 8 10 12 14 Moles of bound H + ions per mole of sample Precipitation detected at pH 9.07 (Kinetic point) While chasing equilibrium, all data points fall on the Equilibrium Precipitation Graph Equilibrium Precipitation Graph Kinetic Precipitation Graph

Overview of Chasing Equilibrium CheqSol adds HCl or KOH solution while precipitate is present, and records the rate of pH change* *after waiting until the onset of sustained response This forces the neutral species to cycle between two states: Between these states, a point will be crossed where the concentration of neutral species would be at equilibrium This technique is called Chasing Equilibrium NOTE: CheqSol is short for Ch asing eq uilibrium Sol ubility CheqSol was invented in April 2004 at Sirius. Sirius have a patent for the CheqSol method. supersaturated (excess neutral species in solution) subsaturated (excess undissolved neutral species) / 70

When KOH is added to raise the pH, the solution of Pindolol becomes supersaturated before it precipitates When precipitate first appears there is a good deal of dissolved unionized material B, but it could take hours before it all falls out of solution KOH addition pauses when precipitation has been detected Molecules of B interact to form particles of precipitate, and BH + ions convert to B to replace some the B that was lost. This releases H + ions, and the pH goes down Supersaturated Pindolol / 70 B(aq) + H + B(s) BH + (aq) Solid State Solution CheqSol reports the gradient of this line pH Time (s) 8.6 8.7 8.8 30 40 50 60

Subsaturated Pindolol After recording the linear fit to the gradient, CheqSol adds HCl to lower the pH. Some of the dissolved B converts to BH + in solution The solution becomes subsaturated, i.e. there is an excess of precipitate that could dissolve B dissolves. Some of it converts to BH + in solution. This consumes H + ions, and the pH goes up / 70 pH Time (s) 8.5832 8.5840 8.5848 8.5856 0 25

Chasing Equilibrium continues until a graph like this can be drawn. In the graph above, there are eight crossing points . After collecting a specified number of crossing points, the instrument adjusts pH back to the starting pH to re-dissolve the sample, then cleans the probes for the next experiment. The Crossing Point Graph for Pindolol Black lines and circles - nothing added Blue lines and triangles - KOH added Red lines and triangles - HCl added The system would be at equilibrium at the crossing points / 70 dpH/dt -0.04 0 0.04 36 56 Time (minutes) Subsaturated basic sample is dissolving Supersaturated basic sample is precipitating Crossing points

Calculating the result for Pindolol dpH/dt Concentration (µg/mL) -0.04 0 0.04 0 10 20 80 90 41.32 µg/mL dpH/dt -0.04 0 0.04 6 36 Time (minutes) Subsaturated basic sample is dissolving Supersaturated basic sample is precipitating Crossing points The concentration of unionized species at each point in the Crossing Point Graph is calculated. This requires weight of sample total volume of solution concentrations and volumes of acid and base dispensed pH at each point pK a (s) of sample, both value and type (acid, base)* Gradient vs. concentration is plotted in the graph below The average value of all crossing points is the concentration of the unionized species at equilibrium. This is the Intrinsic Solubility CV shows the quality of assay CV * The procedure is sensitive to errors in pK a – an error of 1 pK a causes an error of 1 logS unit / 70

Pindolol is a chaser Pindolol is a chaser because its kinetic solubility is significantly higher than its Intrinsic solubility The neutral species of Pindolol forms a supersaturated aqueous solution It precipitates slowly, and would take a long time for all the substance to precipitate at a given pH If Pindolol were to pass from the stomach to the upper intestine, it may be expected to form a supersaturated solution before precipitation – this may help to drive absorption. The ratio between the kinetic solubility and intrinsic solubility provides an indication of the supersaturation factor – a useful number for modelling absorption? / 70

Everything a “Chaser”? When we initially discovered CheqSol – we thought every drug would supersaturate to some degree, and therefore “Chase Equilibrium” Our first paper presents 6 chasers: Stuart, M. Box, K. Chasing equilibrium: measuring the intrinsic solubility of weak acids and bases. Anal. Chem. 2005 (77(4)) pp 983-990 / 70

All measurements at 25 º C in aqueous 0.15M KCl solution Results for Propranolol and Famotidine are the mean of 6 (others are mean of 10) Time taken includes dissolution time Solubility of six compounds from our first paper pK a Sample weight (mg) Time taken (min) Kinetic solubility (µg/mL) Intrinsic solubility (µg/mL) 33 45 ± 6 0.9 ± 0.1 0.8 ± 0.2 [1] 43 180 ± 10 50 ± 4 49 ± 2 [1] 79 4600 ± 900 3500 ± 100 3810 ± 20 [3] Lidocaine 7.95 96-280 3.4-24 Ibuprofen 4.35 6.2-51 Diclofenac 3.99 CheqSol Intrinsic solubility (µg/mL) Literature 61 5900 ± 650 740 ± 40 1100 ± 200 [1] 60 120 ± 1 5.3 ± 0.2 5.6 ± 0.3 [2] 60 340 ± 20 81 ± 6 70 ± 20 [1] Propranolol 9.54 10-19 102-123 Warfarin 4.94 10-12 Famotidine 6.77, 11.01 [1] Avdeef, A. Berger, C M. Brownell, C. Pharm. Res. 2000, 17 (10, 85-89 [2] Bergstr öm, C A S. Strafford, M. Lazorova, L. Avdeef, A. Luthman, K. Artusson, P. J. Med. Chem. 2003, 46, 558-570 [3] Powell, M F. in Analytical Profiles of Drug Substances: Florey, K (ed); Academic \Press, San Diego 1986, 15, 761-779 / 70

Everything a “Chaser”? We then discovered some compounds that did not follow the “Chasing Equilibrium” process. We named these compounds “Non-Chasers” Example – Verapamil ( tertiary amine – a base with pK a of 8.72 @25°C, 0.15M ionic strength) BH + soluble B insoluble / 70

Verapamil Bjerrum Curve For Verapamil, all points collected, including the kinetic solubility, fall on the Precipitation Bjerrum Graph After titrating with base, the instrument adds acid to check that the points are still on the Precipitation Bjerrum Graph. If they are, then … Kinetic solubility = Intrinsic solubility = the solubility value required to fit a precipitation curve to the data points Assay took just 19 minutes to measure kinetic and Intrinsic solubility Precipitation detected at pH 7.82 (Kinetic point) Moles of bound H + ions per mole of sample pH 0.0 0.5 1.0 0 2 4 6 8 10 12 14 All data points collected after precipitation fall on the same Precipitation Bjerrum Graph as the Kinetic Point / 70

Other Non-Chasers Chlorpromazine, Imipramine, Quinine, Amitryptyline, Diphenhydramine, Nortriptyline, Desipramine, Diltiazem, Deprenyl . These compounds would not supersaturate and therefore would precipitate as soon as they exceed their solubility limit. Questions raised: What implications does this have for oral absorption? What determines the degree to which a compound will supersaturate or not? Can we predict supersaturation behaviour from structure? Do chasers precipitate and dissolve at equal rates? / 70

Can we predict whether a sample is a non-chaser? Imipramine Non-chaser Amitryptyline Non-chaser Chlorpromazine Non-chaser Desipramine Non-chaser Nortriptyline Non-chaser Maprotiline Chaser Secondary and tertiary amines with logP > 4. Chlorprothixene Non-chaser converts to chaser Similar structures, but maprotiline contains a -CH 2 -CH 2 - bridge. Non-chaser Trimipramine / 70

More non-chasers.….. Amiodarone Verapamil Quinine Diltiazem Deprenyl / 70

..….. and some chasers Terfenadine Nadolol Loperamide Metoclopramide Amodiaquin Pyrimethamine / 70

Investigating precipitation and dissolution behaviour Piroxicam Sulfamerazine Supersaturated acidic sample Subsaturated acidic sample Subsaturated acidic sample Supersaturated acidic sample Most of the early compounds we investigated show a tight symmetry. / 70

Investigating precipitation and dissolution behaviour Papaverine Furosemide Supersaturated acidic sample Subsaturated acidic sample Subsaturated basic sample Supersaturated basic sample Some compounds show a clear offset! / 70

Using the Precipitation Rate graph to investigate ~100 ionisable drugs, we have found that there appears to be four classes of behaviour. The Four Class Model / 70 Slow Precipitator Fast Precipitator Slow Dissolver “ Chasers” Slow rate for both precipitation and dissolution Examples: Ibuprofen, Benzocaine, Benzthiazide. “ Non-Chasers” Fast rate of precipitation, slow rate of dissolving Examples: Nortriptyline, Amitryptyline, Imipramine. Fast Dissolver “ Super Dissolvers” Slow rate of precipitation, Fast rate of dissolving Examples: Tolmetin, Papaverine, Chlorzoxazone. “ Ghosts” Fast rate for precipitation and dissolution Examples: None yet discovered.

CheqSol Technology A weak base might dissolve fully in the stomach but precipitate on entering the high pH environment of the upper intestinal tract. Can the patterns we observe in CheqSol be used to identify which formulation/delivery methods can be used to improve bioavailability? Does the supersaturation exhibited by “chasers” mean that the bioavailability is already enhanced over what the thermodynamic properties imply, and thus further formulation/delivery work is unwarranted? Do non-chasers fall out of solution as amorphous material whereas chasers produce crystalline precipitate? Amorphous materials are amenable to solid state dispersion nanoparticle delivery methods. Alternatively, could a formulation technique be used to keep a supersaturated sample in a supersaturated state for longer than expected? Are the properties we observe inherent to the compound, or can they be changed by the use of excipients, milling techniques etc.? CheqSol is a unique tool for investigating the precipitation characteristics of a drug. / 70

Validation of CheqSol solubility method CheqSol vs. Shake-flask results for compounds 1– 14 Six replicate shake-flask experiments compared with six CheqSol experiments for each compound Box, K J. Völgyi, G. Baka, E. Stuart, M. Takács-Novák, K. Comer, J E A. J. Pharm. Sci. 2006, in press / 70 19 compounds in this group – see next slide

CheqSol validation 19 compounds with low solubility Good correlation between CheqSol and Shake-flask results / 70

Conclusion Sirius are experts in physicochemical measurement, serving a global market. Sirius provide several turn-key tools for measuring important physicochemical parameters - pK a , LogP/D and Solubility. Our unique CheqSol assay measures solubility AND supersaturation Our unique CheqSol assay provides information on dissolution AND precipitation rates Highly automated instrumentation is available We are always working to improve and update our instruments and software through innovative research, publications and and collaboration. / 70

Sources of Further Information Website: www.sirius-analytical.com Overview of Sirius, our products and our technologies Detailed list of all our literature publications Brochure Downloads Our recent publications: Stuart, M. Box, K. Chasing equilibrium: measuring the intrinsic solubility of weak acids and bases . Anal. Chem. 2005 , 77(4), 983-990 Box, K J. Völgyi, G. Baka, E. Stuart, M. Takács-Novák, K. Comer, J E A. Equilibrium vs. kinetic measurements of aqueous solubility, and the ability of compounds to supersaturate in solution - a validation study . J. Pharm. Sci. 2006 , 95, 1298-1307. Sköld, C. Winiwarter, S. Johan Wernevik, J. Bergström, F. Engström, L. Allen, R. Box, K. Comer, J. Mole, J. Hallberg, A. Lennernäs, H. Lundstedt, T. Ungell, A-L. Karlén, A. Presentation of a Structurally Diverse and Commercially Available Drug Data Set for Correlation and Benchmarking Studies . J. Med. Chem. 2006, 49(23), 6660-6671 Llinàs, A. Burley, J C. Box, K J. Glen, R C. Goodman, J M. J. Diclofenac Solubility: Independent Determination of the Intrinsic Solubility of Three Crystal Forms . Med. Chem.; 2007, 50 (5), 979-983 Collaborative research with the University of Cambridge Llinàs, A. Box, K J. Burley, J C.Glen, R C. Goodman, J M.J. A new method for the reproducible generation of polymorphs: two forms of Sulindac with very different solubilities. J. Applied Crystallography, 2007, 40(2), 379-381. Collaboration with the University of Cambridge. / 70

Measuring pKas, logP and Solubility by Automated titration

More Related Content

What's hot (20)

Viewers also liked (10)

Similar to Measuring pKas, logP and Solubility by Automated titration (20)

More from Jon Mole (7)

Recently uploaded (20)

Measuring pKas, logP and Solubility by Automated titration

Editor's Notes