(583e) Investigation of Phase Separation Phenomena in a Supersaturated Ternary Amorphous Solid Dispersion

Kelly, C., Bernal Institute, University of Limerick
Albadarin, A., University of Limerick
Walker, G., Bernal Institute, University of Limerick

of phase separation phenomena in a

ternary amorphous solid dispersion


Catherine B. Kelly1*, Ahmad B. Albadarin1& Gavin M. Walker1


1Synthesis and Solid State Pharmaceutical
Cluster (SSPC), Bernal Institute,

University of Limerick,
Limerick, Ireland
+353-(0)-61-237732, E-mail:
Catherine.Kelly@ul.ie Abstract

Many new chemical entities (NCEs) are failing to progress through
drug-development pipelines due to a display of insufficient bioavailability, stemming
from the extremely low aqueous solubility properties of many of these
materials. The overwhelming majority of these NCEs are small-molecule,
crystalline solids so lend themselves to formulation in the higher solubility
but thermodynamically unstable amorphous form. A common approach to the
formulation of amorphous APIs is through the use of an amorphous solid
dispersion. In order to stabilise the thermodynamically unstable amorphous form
an excipient to which the API exhibits an affinity through molecular
interactions is added during formulation.

Recent publications have suggested that even in cases in which
interactions between an API and a biocompatible polymer has been detected, in
order to produce an acceptable solid dosage form the amorphous API could be
present at a loading above that which is thermodynamically stable. At the
required higher drug loadings successful long-term stability appears to be achieved
through a combination of thermodynamic and kinetic mechanisms.

Additionally, the action of the excipients in an amorphous solid
dispersion is twofold. As well as the stabilisation of the amorphous form in
the solid state, the excipients should also be capable of maintaining the
supersaturated concentration of the API which occurs during dissolution due to
the presence of the amorphous form. Due to the dual-action properties required
of the polymeric excipient the use of two excipients is proposed, with this study
investigating the first mechanism, namely the stabilisation of the amorphous
form in the solid state.

Production of the drug-polymer temperature-composition phase diagram

Itraconazole (ITR) was used as a model poorly aqueous soluble drug and
Soluplus and hydroxypropyl methylcellulose phthalate (HPMCP) were investigated
as compatible excipients for the formulation of a ternary amorphous solid
dispersion. The value of the Flory-Huggins interaction parameter and its
dependence on temperature was determined for Itraconazole with each polymer as
well as a 1:1 blend of both polymers using the dissolution end point method.
This method determines the liquid-solid boundary in the API-polymer binary
system by measuring the endset of melting during a heating ramp in a
differential scanning calorimeter (DSC). The melting point value gained from
this method was further investigated using a 16-hour annealing method which
indicated that the liquid-solid curve had been underestimated for both
polymers. The annealing method involved the use of hyper DSC and a heating rate
of 100°C/min to search
for the presence of residual crystalline API following a 16 hour hold at or
below the temperature gained using the dissolution end point method.

It is known that the dissolution end point method of determining the
liquid-solid curve in high viscosity systems such as these can overestimate the
drug melting temperature, however the greater accuracy associated with the
annealing method should be considered against the much greater period of time
required to gain a phase diagram data point.

Investigation of phase separation

Following production of the three phase diagrams a number of amorphous
solid dispersions with a range of drug-polymer compositions were produced
through ball milling and a subsequent melt-quench. Hyper DSC was used to
measure the temperature and width of the glass transition and confirm the
absence of crystalline API. Itraconazole is also known to form a liquid-crystal
state but the thermal events associated with this state were not apparent in
the any of the thermograms of the solid dispersions. The dispersions were
stored at 100°C which is
within the region of the phase diagram where spontaneous spinodal decomposition
should occur (Figure 1). After one week phase separation had occurred in all
samples but the development of two distinct demixing mechanisms was apparent.
At higher drug loadings demixing occurred by a release of the API by
recrystallisation. At lower drug loadings the amorphicity of the formulations
was maintained, however two distinct amorphous phases appeared, identified by
two separate glass transition events. The change in demixing from
recrystallisation to phase separation was observed in all of the three systems
examined but shifted in drug-polymer ratio depending on the polymer(s)

Figure 1: Phase diagram of the ITR-HPMCP-Soluplus system indicating the
boundaries identified using the dissolution end point (full line) and annealing
(dashed line) methods and the points at which phase separation was investigated
(crosses). The upper boundary marks that between liquid and solid API while the
lower is the spinodal curve.