Accepted Manuscript

Title: Amorphous is not always betterA dissolution study on

solid state forms of carbamazepine

Authors: Linda G. Jensen, Frederik B. Skautrup, Anette

Mullertz,
Bertil Abrahamsson, Thomas Rades, Petra A.
Priemel

PII: S0378-5173(17)30161-8
DOI: http://dx.doi.org/doi:10.1016/j.ijpharm.2017.02.062
Reference: IJP 16466

To appear in: International Journal of Pharmaceutics

Received date: 14-1-2017

Revised date: 21-2-2017
Accepted date: 22-2-2017

Please cite this article as: Jensen, Linda G., Skautrup, Frederik B., Mullertz,
Anette,
Abrahamsson, Bertil, Rades, Thomas, Priemel, Petra A., Amorphous is not always
betterA dissolution study on solid state forms of carbamazepine.International Journal
of Pharmaceutics http://dx.doi.org/10.1016/j.ijpharm.2017.02.062

This is a PDF file of an unedited manuscript that has been accepted for publication.
As a service to our customers we are providing this early version of the manuscript.
The manuscript will undergo copyediting, typesetting, and review of the resulting proof
before it is published in its final form. Please note that during the production process
errors may be discovered which could affect the content, and all legal disclaimers that
apply to the journal pertain.
Amorphous is not always better a dissolution study on solid state forms of carbamazepine

Linda G Jensen1, Frederik B Skautrup1, Anette Mllertz1,2, Bertil Abrahamsson3, Thomas Rades1*,

Petra A Priemel1
1
University of Copenhagen, School of Pharmaceutical Sciences Denmark,
2
Bioneer:Farma, Faculty of Health and Medical Sciences, University of Copenhagen, Denmark
3
AstraZeneca R&D Mlndal, Pepparedsleden 1, SE-431 83 Mlndal, Sweden

*
Corresponding author:

Thomas Rades
Department of Pharmacy
Faculty of Health and Medical Sciences
University of Copenhagen
Universitetsparken 2
2100 Kbenhavn
DENMARK

Telephone +4535336032
e-mail: thomas.rades@sund.ku.dk

Graphical abstract

1
Abstract

Poor aqueous solubility is a major concern for many new drugs. One possibility to overcome this
issue is to formulate the drug as a high energy form, i.e. a metastable polymorph, an amorphous
neat drug or a glass solution with polymers. In this study the dissolution properties of different solid
state forms of carbamazepine, crystalline or amorphous drug, with or without either
polyvinylpyrrolidone (PVP) or hydroxypropylmethylcellulose (HPMC) and glass solutions of the
drug with both polymers (2:1, 4:1 and 10:1 (w/w) drug-to-polymer ratio) were tested with respect to
their dissolution behaviour in a biorelevant gastric medium (for 30 min) and subsequently in
intestinal conditions (for 2 hours). Carbamazepine form III in the absence of polymer dissolved to a
drug concentration of 540 g/ml, but the concentration decreased after around 70 min due to
precipitation of the dihydrate form, and reached 436 g/ml after 2.5 hours dissolution testing. The
presence of PVP led to a similar dissolution profile with a slightly earlier onset of decrease in drug

2
concentration, while in the presence of HPMC no decline in dissolved drug concentration was
observed. Surprisingly, amorphous carbamazepine did not result in any supersaturation and the drug
concentration was lower than that measured for crystalline carbamazepine. The addition of
polymers further decreased the concentration of dissolved drug (290-310 g/ml, depending on
polymer type and concentration). Amorphous drug converted quickly into the dihydrate form and
thus no supersaturation was achieved. Glass solutions of carbamazepine with PVP reached drug
concentrations between 348 and 408 g/ml after 2.5 hours, i.e. lower than for the crystalline drug,
whilst glass solutions with HPMC reached concentrations similar to the crystalline drug.

Keywords: amorphous, dihydrate, dissolution, supersaturation, carbamazepine

3
1. Introduction

The poor aqueous solubility of many new and existing drugs presents a major formulation challenge

in order to reach suitable drug concentrations at the side of action (Stegemann et al., 2007).

Specifically in the case of oral drug delivery, which is by far the preferred route of drug

administration, dissolution of the drug in the gastrointestinal tract is a prerequisite for drug

absorption. A highly investigated approach to increase the dissolution rate of poorly soluble drugs is

to convert the stable crystalline form of the drug into a high energy form, such as a high energy

(metastable) polymorphic form or an amorphous form (Blagden et al., 2007; Grohganz et al., 2013).

While these approaches increase the apparent solubility and thereby potentially also the

bioavailability of the drug, high energy forms are not thermodynamically stable. Thus they have the

potential to re-crystallise during storage or dissolution to a more stable crystalline polymorph with

an accompanying loss of the dissolution rate and solubility advantage (Priemel et al., 2012;

Savolainen et al., 2009; Tian et al., 2007). One way to overcome this problem is to prepare a glass

solution, in which the drug is molecularly mixed with a polymer (Grohganz et al., 2014). The

dissolution of high energy forms, such as metastable polymorphs, amorphous drugs and glass

solutions, is generally expected to lead to supersaturation of the drug and potentially to subsequent

precipitation (Madsen et al., 2016; Ozaki et al., 2013; Sun and Lee, 2013).

Upon oral administration, most drugs are absorbed in the small intestine. It is therefore important,

especially for amorphous drugs, prone to supersaturation, to understand their behaviour in media

simulating the composition of the gastric and intestinal fluids. The aim of the present study was to

investigate the dissolution properties of different solid state forms of carbamazepine, i.e. of the pure

crystalline form (polymorphic form III, the stable polymorphic form of the drug at room

temperature, but not the stable form in aqueous media at 37 C) or neat amorphous drug (prepared

4
by quench cooling of polymorphic form III), in the presence or absence of either

polyvinylpyrrolidone (PVP) or hydroxypropylmethylcellulose (HPMC) (two commonly used

precipitation inhibitors) and glass solutions of the drug with both polymers (at a 2:1, 4:1 and 10:1

(w/w) drug-to-polymer ratio, prepared by quench cooling) under biorelevant conditions.

2. Materials and Methods

2.1. Materials

Phosphatidyl choline was obtained from Lipoid GmbH (Ludwigshafen, Germany). Hydrochloric

acid 37%, taurocholic acid sodium salt hydrate (>95%), sodium phosphate dibasic, sodium

phosphate monobasic monohydrate, pepsine from porcine gastric mucosa, crude bovine bile and

N,N-dimethylacetamide (DMA) were purchased from Sigma Aldrich (St. Louis, USA). CBZ, PVP

90F and HPMC 615 were purchased from Fagron (Rotterdam, The Netherlands), from BASF

(Ludwigshafen, Germany) and from Cellu ApS (Herlev, Denmark), respectively. Sodium chloride

was purchased from Merck (Whitehouse Station, USA). Methanol (99.9 %) was purchased from

VWR Chemical (Rdovre, Denmark). All chemicals used were of analytical reagent grade or

higher.

2.2. Preparation of samples for dissolution testing

Crystalline CBZ was used as received; the polymorphic form was form III. For amorphous CBZ a

total of 600 50 mg was placed on a tin foil plate. The sample was then melted on a pre-heated

hotplate set to 207 C for 3 min, removed from the hotplate and quench cooled with liquid nitrogen.

Samples were immediately placed over silica and equilibrated at room temperature. Amorphous

CBZ was ground and sieved to a particle size between 180-300 m. For formation of glass

5
solutions, CBZ was pre-mixed with either PVP or HPMC in a 2:1, 4:1 and 10:1 (w/w) ratio and

then treated as described above. All samples were prepared in triplicates.

2.3. DISS Profiler set-up

Dissolution testing was performed in the DISS Profiler with cross section magnets and fiber

optic probes with 2 mm mirror, all from PionInc (Woburn, USA). The fiber optic probes were used

for concentration measurements in the UV range from 305-350 nm. The parameters during the

dissolution studies were set to 180 rpm and 37 C. A standard curve was prepared using an area-

under-the-curve of the second derivative spectral method (Berger et al., 2007; Bijlani et al., 2007;

Bynum et al., 2001; Fagerberg et al., 2010; Madelung et al., 2014) to determine CBZ

concentrations. The samples were tested as powders. Amorphous particles had a size between 180-

300 m and the crystalline CBZ particle size was measured with a Mastersizer (d0.1 3.5 m, d0.5

15.1 m and d0.9 61.7 m).

2.4. Dissolution testing in the DISS

The dissolution behaviour of 20 mg CBZ, either crystalline, amorphous or as glass solution, was

tested in 10 ml Fasted State Simulated Gastric Fluid (FaSSGF, see Table 1) (Vertzoni et al., 2005)

for 30 min. After the dissolution testing the remaining solids were vacuum filtered and then directly

measured afterwards, to avoid artefact creation. The solid state of the filtrates was analysed using

X-ray powder diffractometry (XRPD), see section 2.5. In a second set of dissolution tests, 10 ml of

a conversion buffer (see Table 1) were added after 30 min to the FaSSGF media. The conversion

buffer changed the FaSSGF into a biorelevant intestinal media. Dissolution testing was continued

for another 2 hours and samples were then treated in the same way as described above to gain

information about the solid state form of the remaining solid.

6
2.5. XRPD

XRPD analysis was performed using a XPert PRO X-ray diffractometer from PANalytical

(Almelo, The Netherlands). Samples were measured from 5-35 2 at 45 kV and 40 mA with a

scanning speed of 0.6565 2/min and a step size of 0.01313 2. Data were collected using XPert

Data Collector software from PANalytical.

2.6. Precipitation studies with supersaturated CBZ

CBZ form III was dissolved in DMA at a concentration of 100 mg/ml. The precipitation behaviour

of CBZ was tested by adding 200 L of the CBZ solution to 10 ml biorelevant intestinal media

either containing pre-dissolved HPMC or PVP or without polymer. The amount of polymer that was

added to the medium was chosen such that 2:1, 4:1 and 10:1 (w/w) of drug-to-polymer ratios were

reached, respectively. Precipitation was measured with the DISS Profiler described in section

2.3 using a 10 mm mirror, a stirring rate of 100 rpm and a temperature of 37 C. The light

obscuration (obs) was measured at a wavelength of 500 nm every second and the time to

precipitation was determined as the time where an increase in obs was seen (using a linear

regression of the measuring points showing increasing obs to calculate the intercept with the time

axis, which was then used as the time to precipitation).

3. Results and discussion

3.1. Dissolution behaviour of crystalline CBZ in FaSSGF and biorelevant intestinal media in

the absence of polymers

7
When crystalline CBZ form III was added to biorelevant gastric media, within 10 min a plateau was

reached and maintained until the end of the gastric dissolution step (30 min) with a concentration of

475 12 g/ml. Since drug dissolution was not complete in the medium, XRPD analysis of the

remaining solid could be performed and revealed that CBZ remained in its original polymorphic

form, form III (see Supplementary Material). In the second set of experiments, the conversion

buffer was added to the FaSSGF after 30 min dissolution testing in biorelevant gastric medium

converting the medium into a biorelevant intestinal medium. Through this step the volume in the

measuring vessel doubled and hence concentration of dissolved CBZ halved at 30 min in the

dissolution curves (see Figure 1) and CBZ re-dissolved.

In the absence of polymer, CBZ reached a concentration of 540 g/ml after 20 min of dissolution in

the intestinal medium (50 min total dissolution time). However, after 67min (total dissolution time)

the drug concentration started to decline and at the end of the experiment, after 2h in intestinal

biorelevant medium, a concentration of 436 17 g/ml was measured. XRPD measurements of the

solid phase at the end of the experiment revealed that CBZ form III changed to the crystalline CBZ

dihydrate form during exposure to the intestinal dissolution medium. The dihydrate form has a

lower aqueous solubility than CBZ form III and is the most stable polymorph in aqueous conditions

(Tian et al., 2006b). Hence, the dissolution of the crystalline CBZ form III leads to supersaturation

with regards to the dihydrate. The dihydrate form nucleates and precipitates leading to a decrease in

the measured concentration. This solution mediated transformation of CBZ form III to the dihydrate

has previously been reported (Murphy et al., 2002).

3.2. Dissolution behaviour of crystalline CBZ in FaSSGF and biorelevant intestinal media in

the presence of pre-dissolved polymers

8
Dissolution experiments in biorelevant gastric media were next performed in the presence of PVP

or HPMC, pre-dissolved in the gastric medium. The amount of polymer that was added to the

medium was chosen such that 2:1, 4:1 and 10:1 (w/w) of drug-to-polymer ratios were reached,

respectively. No polymer was added to the conversion buffer in order to keep the ratio of drug-to-

polymer constant.

At drug-to-polymer ratios of 2:1, 4:1 and 10:1 (w/w) the concentration of dissolved drug reached at

the end of the gastric step was 485 13, 481 10 and 447 11 g/ml for PVP and 480 10, 458

4 and 433 17 g/ml for HPMC, respectively. The dissolution kinetics were similar to those

observed for pure CBZ (see Figure 2a and d). As was the case in the absence of polymers, XRPD

confirmed that undissolved CBZ remained to be present as polymorphic form III (see

Supplementary Material).

When the conversion buffer was added to the FaSSGF containing polymers after 30 min of

dissolution testing, a decrease in CBZ concentration at around 50 min for all CBZ-PVP ratios was

observed. At the end of the intestinal dissolution step, CBZ-PVP samples reached concentrations of

around 400 g/ml and the solid phase was analysed by XRPD. Only reflections of the dihydrate

form of CBZ were present in the diffractograms (see Supplementary Material). In contrast, when

HPMC was present in the dissolution medium, no decline in the concentration of dissolved drug

was observed and the concentration remained around 540 g/ml until the end of the intestinal

dissolution step (Figure 2a). XRPD measurements after the dissolution test confirmed that CBZ

form III did not convert to the dihydrate form at all tested concentrations of HPMC.

Previous studies have found contradicting results with regards to the inhibition of the conversion of

CBZ form III to the dihydrate by PVP. Single CBZ crystals were mounted on microscope stubs and

immersed either in water or in a 1% (w/v) aqueous PVP solution (Tian et al., 2006). At pre-

9
determined time points samples were imaged by scanning electron microscopy (SEM). In the

absence of polymer, dihydrate needles grew on the crystal surface. However, when PVP or

hydroxypropyl cellulose were present, the authors observed the dissolution of the drug without any

dihydrate formation. Sodium carboxymethylcellulose and polyethylene glycol 6000 delayed the

hydrate formation, but could not prevent it. HPMC was not among the investigated polymers in that

study. In another study the transformation of CBZ form III in a slurry was investigated using

Raman spectroscopy (Gift et al., 2008). In this paper the authors investigated the effect of added

polymers on nucleation and precipitation using a light scattering probe and on crystal growth with

microscopic measurements. A supersaturated solution of CBZ was created via cooling a saturated

solution from 60 to 25 C. The addition of PVP showed no inhibitory effect on the nucleation

process to the dihydrate. The lack of an inhibitory effect of PVP on the nucleation of CBZ dihydrate

was confirmed in the current study, both in dissolution (see above) and precipitation studies (see

below). For the latter, CBZ was dissolved in DMA and added to the intestinal medium in the

absence and presence of pre-dissolved polymers. The concentration of CBZ was chosen such that it

was supersaturated in the intestinal medium and to a degree that nucleation and precipitation

occurred after a few seconds in the absence of polymer.

In the presence of PVP, unexpectedly the time to precipitation of CBZ was even getting slightly

shorter than in the absence of polymer with decreasing drug-to-polymer ratio (Figure 3). This

finding is supported by the slightly earlier decrease in dissolved drug concentration in the intestinal

step of the dissolution study in the presence of PVP (Figure 2d).

In contrast, in the presence of pre-dissolved HPMC the time to precipitation of CBZ was strongly

increased in the presence of polymer with decreasing drug-to-polymer ratio.

10
The influence of the presence of excipients (dissolved in the media) on the polymorphic form and

dissolution behaviour of CBZ during intrinsic dissolution testing was investigated (Tian et al.,

2007). The excipients HPMC and PEG were found to inhibit (HPMC) or slow down (PEG) the

conversion to the dihydrate, whereas this conversion happened very fast in water. CBZ form III had

a higher dissolution rate when the dissolution study was performed in water than when it was

performed in PEG or HPMC solution. The fast dissolution of CBZ in the absence of polymers (i.e.

when conversion to the dihydrate takes place quickly) was unexpected as the dihydrate has a lower

aqueous solubility than CBZ form III. The measured dissolution rates could be explained by

visualizing the drug compacts during and after dissolution using SEM and investigating the solid

state form after dissolution testing with XRPD. Compacts containing only CBZ form III converted

to the dihydrate form. After 20 min of intrinsic dissolution testing needle formation could be

visualised on the compact surfaces. The conversion to the dihydrate form was slowed down by the

PEG solution and needles were observed only after 100 min. In the presence of HPMC no needles

were observed on the compact surface at all. Dissolution testing was performed in compacted discs

which are used in intrinsic dissolution testing to ensure a constant surface area. The conversion of

CBZ form III to the dihydrate in needle form substantially increased the surface area of the compact

and hence increased the dissolution rate. In the current study powder dissolution was performed in a

set-up where the surface area does not play such a crucial role as in intrinsic dissolution testing.

Therefore the inhibition of the dihydrate formation of CBZ by HPMC had a positive effect on the

concentration of the dissolved drug (see Figure 2a).

3.3. Dissolution behaviour of quench cooled CBZ in the absence of polymers in FaSSGF and

biorelevant intestinal media

11
The main purpose of using an amorphous form of a drug, either as a neat amorphous drug or with a

polymer as a glass solution, is to increase the apparent solubility and the dissolution rate (Grohganz

et al., 2014). This creates supersaturation with regards to the most stable crystalline form. However,

supersaturation is thermodynamically not stable and will eventually lead to precipitation.

Precipitation from supersaturated solutions has been previously observed for example for

indomethacin, naproxen, piroxicam, griseofulvin, danazole, felodipine and itraconazole (Ozaki et

al., 2013; Sarode et al., 2013; Sun and Lee, 2013).

Surprisingly, in the current study the amorphous form of CBZ (confirmed by XRPD measurements,

data not shown) reached lower concentrations compared to CBZ form III throughout dissolution in

FaSSGF (see green and pink dissolution curves in Figure 2). XRPD measurements on the remaining

solid confirmed that the amorphous drug had converted to the dihydrate form of CBZ (see

Supplementary Material). The dissolution data suggests that this conversion had happened in the

solid state of the drug, before dissolution, as supersaturation could not be reached. The

concentration of the dissolved drug in FaSSGF reached 311 9 g/ml with quench cooled

amorphous CBZ as the starting solid material. When the FaSSGF medium was converted to the

intestinal medium the drug concentration dropped due to the increase in volume as seen before for

crystalline CBZ. The drug re-dissolved after the addition of the conversion buffer and reached a

plateau, at the end of the intestinal dissolution step at a drug concentration of 337 17 g/ml. Thus,

in the biorelevant intestinal medium no supersaturation was created either.

Savolainen et al. studied the dissolution behaviour of quench cooled CBZ and CBZ dihydrate in a

flow through cell equipped with a Raman probe (Savolainen et al., 2009). The Raman data revealed

that amorphous CBZ converted to CBZ form I, which has a higher solubility than form III, and then

to the dihydrate. However, in the dissolution testing the amorphous samples reached a lower

concentration than the CBZ dihydrate samples. The authors explained this observation by

12
differences in the wetting characteristics and specific surface areas of the samples. The crystalline

samples were compressed from powder whereas the amorphous samples were quench cooled

directly into a mold resulting in one solid block (Savolainen et al., 2009). Whilst the slower

dissolution rate of amorphous samples compared to the crystalline form of the drug is in good

accordance to our findings, the explanation has to be a different one in the current study. The slower

dissolution rate of the quench cooled sample here is more likely due to the fast and direct

conversion of undissolved amorphous drug to the dihydrate of CBZ.

3.3. Dissolution behaviour of quench cooled CBZ in the presence of pre-dissolved polymers in

FaSSGF and biorelevant intestinal media

When the dissolution experiment in FaSSGF was performed in the presence of pre-dissolved PVP

or HPMC at drug-to-polymer ratios of 2:1, 4:1 and 10:1 (w/w) respectively, the concentrations of

dissolved CBZ reached at the end of the gastric step were 292 15, 287 11 and 284 18 g/ml

for PVP and 274 20, 270 17 and 269 13 g/ml for HPMC, respectively (black, red and blue

curves in Figure 2b and e). After the addition of the conversion buffer, amorphous CBZ in the

presence of both polymers re-dissolved slower than amorphous CBZ without polymers. Samples

reached 310 16, 296 17 and 290 14 g/ml in the presence of HPMC and 298 14, 293 12

and 296 5 g/ml in the presence of PVP (black, red and blue curves in Figure 2b and e).

Polymers are often used to prevent or slow down crystallisation of supersaturated drug. As shown

in Figure 3 this was indeed the case for HPMC, but not for PVP. However, in the dissolution study

presence of pre-dissolved polymer in case of both polymers led to a lower dissolved drug

concentration than in the absence of polymer. This may be explained by the fact that in the

precipitation study (Figure 3) the drug was added in a dissolved form whereas in the dissolution

study (Figure 2) the drug was initially present in the solid form not reaching supersaturation.

13
Amorphous CBZ particles may be better wetted in the presence of polymer leading to a faster

conversion to the dihydrate of CBZ.

The drug concentration measured for samples in the presence of both polymers was still slightly

increasing at the end of the dissolution testing (150 min). If the dissolution testing would have been

carried on the concentration measured for quench cooled CBZ without polymers may have been

reached.

3.4. Dissolution behaviour of quench cooled CBZ glass solutions in FaSSGF and biorelevant

intestinal media

Quench cooled CBZ had a lower dissolution rate than crystalline CBZ form III even in the presence

of pre-dissolved polymers. Therefore glass solutions in the same drug-to-polymer ratios 2:1, 4:1 and

10:1 (w/w) were tested to investigate if an intimate contact between the drug and the polymer

molecules would prevent dihydrate formation and thus lead to supersaturation. The amorphous

nature of the glass solutions has been confirmed by XRPD measurements (data not shown).

Glass solutions of CBZ and either polymer reached a higher concentration of dissolved drug than

quench cooled CBZ but a lower concentration than CBZ form III in the gastric step of the

dissolution testing. After addition of the conversion buffer glass solutions of CBZ and PVP reached

a concentration plateau of dissolved CBZ after 20 min of dissolution in the intestinal medium (50

min total dissolution time). The concentration of this plateau was 408 9, 374 9 and 348 13

g/ml for the 2:1, 4:1 and 10:1 (w/w) drug-to-polymer ratios and dependent on the CBZ-PVP ratio.

The glass solution with the highest PVP content reached the highest drug concentration (see blue,

red and black curve in Figure 2f).

In the presence of HPMC the 2:1 and the 4:1 drug-to-polymer ratio had the same dissolution

kinetics (see blue and red curve in Figure 2c). At the 10:1 drug-to-polymer ratio the drug dissolved

14
slower, (see black curve in Figure 2c). At the end of the gastric step a CBZ concentration of 360

6, 373 7 and 350 4 g/ml for the 2:1, 4:1 and 10:1 (w/w) drug-to-polymer ratios was reached.

Directly after the conversion to the intestinal medium CBZ re-dissolved in the same manner as in

the gastric step. At approximately 50 min, dissolution slowed down but continued steadily until the

end of the experiment reaching almost the concentration of dissolved drug from crystalline CBZ.

Diffractograms contained only dihydrate reflections both after gastric and intestinal dissolution

steps (see Supplementary Material).

Whilst the glass solutions of the respective drug-to-polymer ratios performed better than amorphous

drug with pre-dissolved polymer (see Figure 2b and e) and also better than quench cooled

amorphous CBZ alone (see green curve in Figure 2c and f), none of the amorphous forms reached

supersaturation and the area under the dissolution curve was the highest for crystalline CBZ form

III.

4. Conclusions

It is generally expected that amorphous forms of a drug, either neat amorphous drug or a glass

solution of the drug and a polymer, result in a higher apparent solubility, a faster dissolution of the

drug and the creation of supersaturation. The current study, however, has shown that this is not

always the case. Amorphous, quench cooled CBZ with and without pre-dissolved polymers (PVP or

HPMC) and glass solutions of CBZ with PVP or HPMC reached a lower drug concentration in

dissolution testing than crystalline CBZ form III. While amorphous neat CBZ with and without pre-

dissolved polymers, glass solutions of CBZ and CBZ form III all converted to the dihydrate form of

CBZ, this happened faster for the amorphous forms than for the crystalline form. In the presence of

pre-dissolved HPMC, but not of pre-dissolved PVP, the conversion from CBZ form III to the

dihydrate form was prevented, supersaturation with regards to the dihydrate form was maintained

15
and hence the highest concentration during dissolution testing in this study was measured. This

study highlights that formulation approaches to enhance the solubility of a drug have to be chosen

carefully and that there is no one fits all formulation strategy.

Acknowledgements

The authors acknowledge funding from Astra Zeneca and the University of Copenhagen for the

PhD project of Linda Grn Jensen.

References

Berger, C.M., Tsinman, O.,Voloboy, D., Lipp, D., Stones, S., Avdeef, A., 2007. Technical note: miniaturized
intrinsic dissolution rate (Mini-IDR (TM)) measurement of griseofulvin and carbamazepine. Dissolut.
Technol. 14, 3941.
Bijlani, V., Yuonayel, D.,Katpally, S.,Chukwumezie, B.N., Adeyeye, M.C., 2007. Monitoring ibuprofen release
from multiparticulates: in situ fiber-optic technique versus the HPLC method: a technical note. AAPS
PharmSciTech, 8, E9E12.
Blagden, N., de Matas, M., Gavan, P.T., York, P., 2007. Crystal engineering of active pharmaceutical
ingredients to improve solubility and dissolution rates. Advanced Drug Delivery Reviews 59, 617-630.
Bynum, K., Kassis, K.R.A., Pocreva, J., Gehriein, L., Cheng, F., Palermo,P., 2001. Analytical Performance of a
Fiber Optic Probe Dissolution System. Dissolut. Technol. 8, 1322.
Fagerberg, J.H., Tsinman, O., Sun, N., Tsinman, K., Avdeef, A., Bergstrm, C.A.S., 2010. Dissolution rate and
apparent solubility of poorly soluble drugs in biorelevant dissolution media. Molecular Pharmaceutics, 7,
14191430
Gift, A.D., Luner, P.E., Luedeman, L., Taylor, L.S., 2008. Influence of polymeric excipients on crystal hydrate
formation kinetics in aqueous slurries. J. Pharm. Sci. 97, 5198-5211.
Grohganz, H., Lbmann, K., Priemel, P., Tarp Jensen, K., Graeser, K., Strachan, C., Rades, T., 2013.
Amorphous drugs and dosage forms. Journal of Drug Delivery Science and Technology 23, 403-408.
Grohganz, H., Priemel, P.A., Lbmann, K., Nielsen, L.H., Laitinen, R., Mullertz, A., Van Den Mooter, G.,
Rades, T., 2014. Refining stability and dissolution rate of amorphous drug formulations. Expert Opinion on
Drug Delivery 11, 977-989.
Madelung, P., Ostergaard, J., Bertelsen, P., Jrgensen, E.V., Jacobsen, J., Mllertz, A., 2014. Impact of
sodium dodecyl sulphate on the dissolution of poorly soluble drug into biorelevant medium from drug-
surfactant discs. Int. J. Pharm. 467, 1-8.
Madsen, C.M., Boyd, B., Rades, T., Mllertz, A., 2016. Supersaturation of zafirlukast in fasted and fed state
intestinal media with and without precipitation inhibitors. Eur. J. Pharm. Sci. 91, 31-39.
Murphy, D., Rodrguez-Cintrn, F., Langevin, B., Kelly, R.C., Rodrguez-Hornedo, N., 2002. Solution-mediated
phase transformation of anhydrous to dihydrate carbamazepine and the effect of lattice disorder. Int. J.
Pharm. 246, 121-134.

16
Ozaki, S., Kushida, I., Yamashita, T., Hasebe, T., Shirai, O., Kano, K., 2013. Inhibition of crystal nucleation and
growth by water-soluble polymers and its impact on the supersaturation profiles of amorphous drugs. J.
Pharm. Sci. 102, 2273-2281.
Priemel, P.A., Grohganz, H., Gordon, K.C., Rades, T., Strachan, C.J., 2012. The impact of surface- and nano-
crystallisation on the detected amorphous content and the dissolution behaviour of amorphous
indomethacin. Eur. J. Pharm. Biopharm. 82, 187-193.
Sarode, A.L., Sandhu, H., Shah, N., Malick, W., Zia, H., 2013. Hot melt extrusion (HME) for amorphous solid
dispersions: Predictive tools for processing and impact of drug-polymer interactions on supersaturation.
Eur. J. Pharm. Sci. 48, 371-384.
Savolainen, M., Kogermann, K., Heinz, A., Aaltonen, J., Peltonen, L., Strachan, C., Yliruusi, J., 2009. Better
understanding of dissolution behaviour of amorphous drugs by in situ solid-state analysis using Raman
spectroscopy. Eur. J. Pharm. Biopharm. 71, 71-79.
Stegemann, S., Leveiller, F., Franchi, D., de Jong, H., Lindn, H., 2007. When poor solubility becomes an
issue: From early stage to proof of concept. Eur. J. Pharm. Sci. 31, 249-261.
Sun, D.D., Lee, P.I., 2013. Evolution of supersaturation of amorphous pharmaceuticals: The effect of rate of
supersaturation generation. Molecular Pharmaceutics 10, 4330-4346.
Tian, F., Sandler, N., Gordon, K.C., McGoverin, C.M., Reay, A., Strachan, C.J., Saville, D.J., Rades, T., 2006.
Visualizing the conversion of carbamazepine in aqueous suspension with and without the presence of
excipients: A single crystal study using SEM and Raman microscopy. Eur. J. Pharm. Biopharm. 64, 326-335.
Tian, F., Sandler, N., Aaltonen, J., Lang, C., Saville, D.J., Gordon, K.C., Strachan, C.J., Rantanen, J., Rades, T.,
2007. Influence of polymorphic form, morphology, and excipient interactions on the dissolution of
carbamazepine compacts. J. Pharm. Sci. 96, 584-594.
Vertzoni, M., Dressman, J., Butler, J., Hempenstall, J., Reppas, C., 2005. Simulation of fasting gastric
conditions and its importance for the in vivo dissolution of lipophilic compounds. Eur. J. Pharm. Biopharm.
60, 413-417.

17
Figure1. Dissolution of CBZ form III in FaSSGF (0-30 min) and biorelevant intestinal (30-150
min) media.

Figure 2. Dissolution of (a) crystalline CBZ III in the presence of HPMC (pre-dissolved in
medium) at a ratio of 10:1, 4:1 and 2:1 (w/w), (b) quench cooled, amorphous CBZ in the
presence of HPMC (pre-dissolved in medium) at a ratio of 10:1, 4:1 and 2:1 (w/w), (c) CBZ
HPMC glass solutions at a ratio of 10:1, 4:1 and 2:1 (w/w), (d) crystalline CBZ III in the
presence of PVP (pre-dissolved in medium) at a ratio of 10:1, 4:1 and 2:1 (w/w), (e) quench
cooled CBZ in the presence of PVP (pre-dissolved in medium) at a ratio of 10:1, 4:1 and 2:1
(w/w) and (f) CBZ PVP glass solutions in a ratio of 10:1, 4:1 and 2:1(w/w).

18
Figure 3. Precipitation of CBZ in the presence of HPMC (black) and PVP (red) from DMA
solution in intestinal media

19
Table 1 Composition of FaSSGF and the bile salt buffer

Compound Concentration Amount in 1000 ml

Pepsin 0.1 mg/ml 0.10 g

Bile (NaTc) 80 M (537.7 g/mol) 0.043 g

FaSSGF
Phospholipids (Lipoid 20 M (786 g/mol, 99%
0.016 g
S PC) pure PC)
Sodium Chloride 34.2 mM (58.4 g/mol) 2.00 g

NaH2PO4 H2O 3.93 mg/ml 0.39 g

Na2HPO4 1.64 mg/ml 0.16 g

Bile salt buffer 5 mM (489,0 g/mol,
Crude bovine bile salts 0.63 g
70.6% pure)
Sodium Chloride 150 mM (58.4 g/mol) 1.55 g
Phospholipids (Lipoid 0.20 g
1.25 mM
S PC)

20