Strategies for

Working with
Poorly Water-
Soluble APIs J A N U A RY 2 0 2 0

Boosting Drug Bioavailabilty through Crystallization Inhibition

Opportunities and Challenges in Softgel Formulation Development

Optimizing Lipid-Based Drug Delivery Systems:

The Influence of Drug Load and Composition

High Throughput Screening for the Discovery of

Self-Emulsifying Drug Delivery Systems (SEDDS)
Frank Romanski and Ronny Hoffmann

Sponsored by

This custom ebook is sponsored by BASF and presented in partnership with Pharmaceutical Technology.
Boosting
Softgel
Boosting Drug Bioavailabilty
Bioavail-
ability
Formulation
SNEDDS SEDDS through Crystallization Inhibition

S P ONS OR E D C ONTE NT

Boosting Drug
Bioavailabilty through
Crystallization
Inhibition
Directly interfering with crystal
nucleation and crystal growth rate
kinetics promotes API bioavailability.

A
core challenge among drug
developers is working with
challenging APIs, which are
increasingly becoming more
hydrophobic and/or with stronger crystal
lattice structures, each ultimately poorly
water soluble.
This summary of a presentation de-
livered by Frank Romanski, PhD, global
technical marketing manager at BASF, at
the 2018 AAPS PharmSci 360 meeting
discusses the use of crystallization inhibi-
tion and softgels as a strategy for solubili-
zation enhancement.
Crystallization inhibition
To improve the dissolution rate of poorly
water-soluble APIs, many pharmaceuti-
cal developers are utilizing softgels as a
method to encapsulate challenging APIs,
which fully solubilize the API in the dos-
age form using a combination of solvents,
surfactants, and polymers. With careful
selection, these excipients may also
inhibit re-crystallization once the drug is
released inside the aqueous environment
of the GI tract.
The mechanism here is that excipients
first increase the concentration above
the maximum achievable saturation
concentration of the drug. Next, careful
selection of surfactants and polymers
act to delay precipitation from supersat-
urated solutions by inhibiting crystalliza-
tion (C ≥ Ceq), a largely kinetic approach.
This is in contrast to simply increasing the
absolute solubility limit. While clearly this
would be effective, it is not practical in
the much larger fluid volumes of the body
in comparison to the excipients present
inside a softgel.
Crystallization inhibitors help maintain a
certain “available” concentration of a drug
within a medium. If the API re-crystallizes
before absorption, its resulting bioavail-
ability declines, thus reducing the drug’s
therapeutic efficacy.
Like solubility, supersaturation is crucial
to a drug’s bioavailability. As shown in
Figure 1, the intraluminal concentration
of a drug is not necessarily limited by its

2 J ANUA RY 2 0 2 0 | P H A R MT E C H SPO N SO R ED CO N TEN T

Boosting
Softgel
Boosting Drug Bioavailabilty
Bioavail-
ability
Formulation
SNEDDS SEDDS through Crystallization Inhibition

Figure 1: Crystallization inhibition: Defining the differences between solu-

bilization enhancement and crystallization inhibition.

Figure 2: Spring and parachute model: Leveraging the absorption

window.

3 J ANUA RY 2 0 2 0 | P H A R MT E C H SPO N SO R ED CO N TEN T

Boosting
Softgel
Bioavail-
ability
Formulation
SNEDDS SEDDS
Figure 3: Danazol – 500 μL in 50 mL pH 2 HCl buffer.
solubility in GI fluids but rather the time
it can be maintained in a supersaturated
state during the absorption window.
Figure 1 shows that adding a primary
solubilizer to a poorly soluble drug in-
creases its concentration, but causes
a rapid crystallization known as the
“spring effect.” Ideally, crystallization in-
hibition results in the “parachute effect”
(Figure 2), where the production of a
very high supersaturated solution slowly
decreases to baseline level leaving a
wide window for absorption.
Several methodologies are employed
to achieve and maintain APIs in a super-
4 J ANUA RY 2 0 2 0 | P H A R MT E C H SPO N SO R ED CO N TEN T
Boosting Drug Bioavailabilty
through Crystallization Inhibition
saturated state in its absorption window.
Among them are surfactants and cyclo-
dextrins, which reduce the degree of su-
persaturation and improve the maximum
concentration of dissolved API, thereby
preventing the drug from seeding small
crystals.
Specifically in softgel formulations, poly-
meric excipients and surfactant precipita-
tion inhibitors are frequently used because
they directly interfere with the nucleation
of crystals and/or crystal growth rate
kinetics. Crystallization-inhibiting polymers
may prevent clusters of drug molecules
from reorganizing themselves into ordered
Boosting
Softgel
Bioavail-
ability
Formulation
SNEDDS SEDDS
Figure 4: Danazol – 500 μL in 50 mL pH 6.8 phosphate buffer.
structures. Inhibition of drug molecule
reorganization is a rate-limiting step in
the two-step nucleation model.
Next, by hindering diffusion of drug
molecules to the crystal nuclei, poly-
mers can also inhibit the crystal growth
that occurs as a result of change in the
absorption layer at the crystal-solution
interface. Because crystallization in-
hibitors are adsorbed onto the crystal
structure, drug molecules are slowed or
no longer incorporated into the crystal
lattice, and the overall rate of crystal
growth decreases. Crystallization inhibi-
5 J ANUA RY 2 0 2 0 | P H A R MT E C H SPO N SO R ED CO N TEN T
Boosting Drug Bioavailabilty
through Crystallization Inhibition
tors may also alter the surface energy
of the crystal face, thereby changing the
level of solvation.
Predicting crystallization
Pharmaceutical scientists often use
technology solutions to generate infor-
mation that will help them to predict
how a softgel formulation will behave
in the human body. The automated Pion
Inform®, for instance, monitors the re-
crystallization rate of polymers and sur-
factants under simulated GI tract condi-
Boosting
Softgel
Boosting Drug Bioavailabilty
Bioavail-
ability
Formulation
SNEDDS SEDDS through Crystallization Inhibition

Figure 5: Kollidon® VA 64: Extending the window further with higher con-
centrations (intestine pH 6.8, Danazol model).

tions. The inForm measures the kinetics
of dissolution, absorption, controlled
supersaturation, and precipitation.
In one series experiment, the inForm
was used to analyze the effects of solubi-
lizers and crystallization inhibitors on the
softgel fill formulations of model com-
pounds danazol, nifedipin, and loratadine
in supersaturated conditions. Danazol,
a poorly water-soluble drug that is often
used for modeling purposes, is not typical-
ly affected by pH conditions; polymorphic
precipitation kinetics characterize nifedip-
ine; and loratadine has a pKa of 4.33, thus
exhibiting the challenge of retaining drug
in solution during the transition from the
stomach to the intestine.
To mimic the average volume of a softgel
fill, the drug formulations evaluated by
inForm totaled 500 μL. Kollisolv® PEG 400
served as the hydrophilic softgel base fill.
The fills also contained solubilizers and
crystallization inhibitors that were miscible
at the 5% level. Their viscosity was suit-
able for softgel manufacturing. Prior to the
tests, they were blended with polyethyl-
ene glycols (PEGs) in the 500 µL softgel
formulation.

6 J ANUA RY 2 0 2 0 | P H A R MT E C H SPO N SO R ED CO N TEN T

Boosting
Softgel
Bioavail-
ability
Formulation
SNEDDS SEDDS
Figure 6: Nifedipine – 500 μL in 50 mL pH 6.8 phosphate buffer.
Each formulation was injected into a
50-mL buffer system at either a pH 2.0
to replicate stomach conditions or pH 6.8
for intestinal conditions. inForm software
tracked each formulation for at least 30
minutes.
With the danazol formulation in a pH 2
hydrochloric acid buffered solution (gas-
tric, acidic conditions), Kolliphor® RH 40
produced a “spring” effect, because the
API’s solubility increased to a very high
level (150 µg per mL) in both stomach and
intestinal conditions, but soon crashed to
near baseline levels. The inForm analysis
also found that Kolliphor® RH 40 and the
7 J ANUA RY 2 0 2 0 | P H A R MT E C H SPO N SO R ED CO N TEN T
Boosting Drug Bioavailabilty
through Crystallization Inhibition
other traditional surfactants tested may
not contribute significantly to crystalliza-
tion kinetics (Figure 3).
Most surprising was the significant
“parachute” effect found using Kollidon®
VA64, a polymer widely used in amor-
phous solid dispersions, but only seldom
in softgels. Kollidon® VA64 was unques-
tionably the most effective, significantly
improving the absorption window, likely
because of its strong affinity with poorly
soluble APIs, an attribute necessary to
inhibit crystallization.
Similar effects were noted under intesti-
nal conditions (Figure 4).
Boosting
Softgel
Bioavail-
ability
Formulation
SNEDDS SEDDS
Figure 7: Loratadine – 500 μL in 50 mL pH 2 HCl buffer.
Figure 5 shows how higher concentra-
tions (5%, 10%, and 20%) of Kollidon®
VA64 can extend the intestinal absorption
window of danazol. Note that in these
examples, the windows are purposely
short to isolate the effects of individual
surfactants and polymers. The final formu-
lations may incorporate combinations of
these ingredients to widen the absorption
window significantly.
In the stomach tests with nifedipine, the
effects of standard PEG, Kollidon® VA64,
and Kollidon® 12PF formulations did not
differ. However, in the intestinal model
(Figure 6), the effects of the three agents
8 J ANUA RY 2 0 2 0 | P H A R MT E C H SPO N SO R ED CO N TEN T
Boosting Drug Bioavailabilty
through Crystallization Inhibition
were very different. The indistinguishable
effects of Kollidon® VA64 and Kollidon® 12
PF suggest that their polyvinylpyrrolidone
(PVP) moiety interacts favorably with the
crystal surface to limit crystal growth.
While most APIs are absorbed in the
intestines, loratadine was fully solubilized
in the stomach (Figure 7). Because of
loratadine’s 4.33 pKa, most of the drug’s
molecules began to crash as soon as they
entered the intestines (Figure 8). The
“parachute” effect was produced by add-
ing Kolliphor® RH40 to the formulation.
Boosting
Softgel
Bioavail-
ability
Formulation
SNEDDS SEDDS
Figure 8: Loratadine – 500 μL in 50 mL pH 6.8 phosphate buffer.
New Solutions
A relatively new solubizer is Soluplus®, a
graft copolymer of PEG comprising poly-
vinyl acetate and polyvinyl caprolactam.
The novel polymer, which has been ex-
tensively used in hot melt extrusion, was
designed to inhibit crystallization through
quasi-complexation. It interacts very favor-
ably with different API molecules and holds
onto them in a colloidal structure, rather
than a polymeric model in which the inter-
action occurs only with the API’s surface.
Soluplus® also has very strong hydrogen
bonding and Van der Waals interactions.
9 J ANUA RY 2 0 2 0 | P H A R MT E C H SPO N SO R ED CO N TEN T
Boosting Drug Bioavailabilty
through Crystallization Inhibition
The inForm test results on Soluplus® at
5% dissolved in PEG 400 and eight other
crystallization-inhibitors are compared in
Figure 9. In the photo at right, the cloudy
nature and blue tint of the Soluplus® solu-
tion indicates the formation of colloidal
structures from the drug’s interactions
with the solubilizer.
The inForm analysis showed that
Soluplus® and several other polymers
such as Kollidon® VA64 can sufficiently
slow down the precipitation process, thus
allowing sufficient absorption of the API
into the bloodstream.
Boosting
Softgel
Boosting Drug Bioavailabilty
Bioavail-
ability
Formulation
SNEDDS SEDDS through Crystallization Inhibition

Figure 9: Soluplus®: API-polymer interactions: Under investigation –

Soluplus® crystallization inhibition or complexation?

Summary Large molecular weight polymers are

Low molecular weight polyvinylpyrrol-
idones (PVPs) such as Kollidon® 12PF and
Kollidon® 17PF are among the polymers
and surfactants that are known to effec-
tively inhibit API crystallization in softgels.
However, newer applications of polymers
such as Kollidon® VA64 provide the strong
affinity with APIs that is necessary to
strongly inhibit crystallization.
Traditional surfactants such as Kolliphor®
RH40 can increase an API’s solubility but
may not contribute significantly to crystal-
lization kinetics.
perhaps the most effective agents to
prevent nucleation and crystal growth.
In order to be fully effective, the polymer
should form specific favorable interactions
with the API. Strong interactions between
the API and the polymer can inhibit the
drug’s precipitation from solution. The
summary of this work is that there is no
one-size-fits-all approach to crystallization
inhibition, but there are many new tools,
such as Kollidon® VA64 and Soluplus® that
can be incorporated into practical dosage
forms like softgels in order to significantly
improve overall bioavailability.

10 J ANUA RY 2 0 2 0 | P H A R MT E C H SPO N SO R ED CO N TEN T

Boosting
Bioavail-
Softgel
SNEDDS SEDDS
Softgel Formulation Development
Formulation
ability

S P ONS OR E D C ONTE NT

Opportunities and
Challenges in
Softgel Formulation
Development

T
he apparent simplicity of
softgel capsules belies their
formulation and technological
sophistication. This summary
of a presentation delivered by Gabriele
Reich, PhD, of the Department of Phar-
maceutical Technology and Biopharma-
ceutics at the University of Heidelberg,
at the 2018 AAPS PharmSci 360 meeting
provides a brief introduction to softgel
manufacturing technology and discusses
the challenges of softgel formulation
design, describing the effect of excipient
quality on gelatin cross-linking and soft
capsule dissolution performance.
Manufacturing process
Manufacture of softgel capsules starts with
the preparation of the gel mass (a heated
mixture of gelatin, water, plasticizer, and
colorant, if used) that is stored in a heated
reservoir and either pumped or transferred
by gravity to the spreader boxes. From the
heated spreader boxes, ribbons are formed
on the surface of the cooling drums, lubri-
cated, and forwarded to the rotary-die
encapsulation unit. In the encapsulation
unit, the fill is injected from the wedge
into the space between the two gelatin
ribbons passing the cavities on the op-
posing die rolls thereby forming the
capsules. Sealing is performed by means
of heat and pressure.
The molecular mechanism underlying
the ribbon formation is a conformational
change of the gelatin molecules from
random coil at elevated temperature into
a three-dimensional elastic network of
partly triple helical gelatin molecules upon
cooling. The gel mass must have the re-
quired viscosity and setting performance to
quickly form elastic ribbons with a defined
thickness and mechanical strength. Sealing
performance strongly depends on the melt-
ing and setting characteristics of the gel
mass. The successful manufacture of me-
chanically stable softgel capsules relies on
the shell microstructure, which is the result
of a complex interaction between gelatin,
water, and plasticizer during encapsulation
and drying.

11 J ANUA RY 2 0 2 0 | P H A R MT E C H SPO N SO R ED CO N TEN T

Boosting
Bioavail-
Softgel
SNEDDS SEDDS
Formulation
ability
Components of the softgel capsule
Softgel capsule shells consist of
pharmaceutical-grade porcine, bovine, or
fish gelatin of different bloom strengths
(a measure of the rigidity of the gelatin
gel), water, and a plasticizer or plasticizer
mixture (e.g., glycerol, sorbitol and relat-
ed products). The plasticizer-to-gelatin ra-
tio determines the capsule strength and
is adjusted and optimized for the desired
capsule size and the intended use of the
product. Moreover, shell/fill interactions
have to be considered. Using regulatory-
approved colorants, softgel capsules
with or without imprinted markings can
be made in numerous colors.
Softgel capsule
formulation challenges
The ultimate goal of a softgel capsule
formulation design is to minimize poten-
tial shell/fill interactions that may affect
the mechanical performance, storage
stability, disintegration and/or the dis-
solution profile (i.e., to reach the equilib-
rium during the manufacturing process).
Liquid fills often include complex lipo-
philic, hydrophilic, and amphiphilic sys-
tems that can potentially interact at the
shell-fill interface. These interactions can
include migration of water, plasticizer, or
both from shell into the fill, migration of
excipients or drug substance from the fill
into the shell, or both. These exchange
processes can cause degradation or
crystallization of the drug substance.
In addition, when the capsules are
stored at high temperature and/or humid-
12 J ANUA RY 2 0 2 0 | P H A R MT E C H SPO N SO R ED CO N TEN T
Softgel Formulation Development
“The ultimate goal of a
softgel capsule formula-
tion design is to minimize
potential shell/fill interac-
tions that may affect the
mechanical performance,
storage stability, disinte-
gration and/or the dissolu-
tion profile.”
ity, gelatin self-crosslinking may occur.
Moreover, crosslinking of the softgel
shell induced by aldehydes that may be
present in the excipients of the fill may
cause disintegration and/or dissolution
problems. Gelatin self-crosslinking involves
the reaction of carboxylic acid and amine
end groups to form an amide bond that
joins two gelatin molecules together. For
softgel capsules with liquid fills containing
ethoxylated excipients (e.g. polyethylene
glycols of various molecular weights,
polysorbates, and others containing certain
levels of aldehydes), aldehyde-induced
crosslinking comprising arginine–arginine,
lysine–lysine, or arginine–lysine residues
in the gelatin may cause the formation of
pellicles (crosslinked gelatin film) on the
inner surface of the softgel shell.
In ethoxylated fill excipients, alde-
hydes can form via the degradation of
polyethylene glycol to peroxide and
hydroperoxide to aldehyde, predomi-
nately formaldehyde. Interestingly, only
the European Pharmacopoeia specifies
Boosting
Bioavail-
Softgel
SNEDDS SEDDS
Softgel Formulation Development
Formulation
ability

Figure 1: Aldehyde / peroxide content of PEG 400 samples.

Figure 2: Case study #1: preliminary results / 3 months data (25/60).

13 J ANUA RY 2 0 2 0 | P H A R MT E C H SPO N SO R ED CO N TEN T

Boosting
Bioavail-
Softgel
SNEDDS SEDDS
Formulation
ability
Figure 3: Case study #1: preliminary results / 3 months data (40/74).
a limit for the formaldehyde content in
polyethylene glycols, whereas the United
States Pharmacopoeia does not. As a re-
sult, pharmacopoeial quality criteria may
not be sufficient to control the aldehyde,
especially the formaldehyde, content in
commonly used ethoxylated liquid-fill
excipients. In fact, in a study published
in 2007, the total aldehyde content of
30 commonly used ethoxylated liquid-fill
excipients was quantified at ppm levels
using a GC–MS method. These results
showed that the total aldehyde content
ranged from less than 5 ppm to approxi-
mately 120 ppm. Actually, the highest
levels were observed in low molecular
weight polyethylene glycols.
Indeed, the case study presented by
Dr. Reich clearly demonstrated a correla-
tion between the aldehyde content (total
aldehydes, formaldehyde and acetalde-
hyde) of PEG 400 fills (see Figure 1) and
the change in the dissolution profiles
of the soft capsules after three months
storage under ICH conditions (e.g., at
14 J ANUA RY 2 0 2 0 | P H A R MT E C H SPO N SO R ED CO N TEN T
Softgel Formulation Development
25°C / 60% rh and and 40°C / 75% rh,
Figure 2 and Figure 3, respectively).
The dissolution profiles of softgel
capsules containing PEG 400 grades
of high purity (Kollisolv®), which were
produced under nitrogen purging and
contained less than 10 ppm total alde-
hydes, remained unchanged upon stor-
age. However, softgel capsules filled
with fresh and stressed Pluriol® E 400
revealed more or less pronounced chang-
es in their dissolution profiles according
to their aldehyde content. This clearly
indicates that the level of impurities in
PEG 400 is crucial for shell cross-linking.
It was also shown that the sensitivity to
cross-linking can be reduced by the shell
formulation design with respect to the
gelatin type and grade and the plasticizer
type and concentration. For example,
softgel capsules with high amounts of
glycerol as plasticizer allowed the per-
meation of oxygen when stored at high
relative humidity, which promoted forma-
tion of aldehydes in the initially low alde-
Boosting Softgel Formulation Development
Softgel
Bioavail- SNEDDS SEDDS
Formulation
ability

hyde polyethylene “Successful development most reactive im-

glycol-containing fill. purity that induces
On the other hand, and manufacture of soft- crosslinking in gela-
softgel capsules gel capsules requires the tin, the use of PEG
made of special- with low aldehyde
grade gelatins in right combination of fill content (<10 ppm)
combination with and shell excipients to has been suggested
glycerol/Polysorb® (e.g., Kollisolv PEG
as plasticizer system minimize cross-linking 400 LA). The ab-
exhibited dissolution issues and hence prod- sence of oxygen by
profiles that were nitrogen purging has
almost superimpos- uct stability problems in also been shown
able over six months terms of dissolution and to be effective in
at 40 °C and 75% rh. minimizing aldehyde
disintegration.” formation and hence
the crosslinking
Conclusions of the gelatin. At the time of use, the
Successful development and manufac- possible degradation of PEG to e.g. alde-
ture of softgel capsules requires the hydes due to elevated temperature and
right combination of fill and shell excipi- exposure to light and/or air during stor-
ents to minimize cross-linking issues age needs to be considered. In addition,
and hence product stability problems in studies have shown that the use of an
terms of dissolution and disintegration. appropriate gelatin grade and the choice
For example, the use of pharmacopoeial- of the plasticizer system are additional
grade (e.g., EP/USP/NF) polyethylene parameters that are important for drug
glycol 400 (PEG 400) may not be suf- product stability.
ficient for the long-term stability of the
drug product. Since formaldehyde is the

15 J ANUA RY 2 0 2 0 | P H A R MT E C H SPO N SO R ED CO N TEN T

Boosting
Bioavail-
Softgel
SNEDDS SEDDS
Optimizing Lipid-Based Drug Delivery
Formulation
ability

S P ONS OR E D C ONTE NT

Optimizing Lipid-Based
Drug Delivery Systems:
The Influence of Drug
Load and Composition
Lipid-based drug delivery systems
can be tailored for individual drug
properties through an in depth
understanding of the mechanisms
of absorption and digestion.

O
ptimizing a lipid-based drug
delivery system requires a
rational approach, taking into
account the type of excipients
used, the solubility of the drug in the excipi-
ents, the effects of digestion within the gut,
and application of in vitro models simulating
the digestion in the human gastrointestinal
tract. Self-nano-emulsifying drug delivery
systems have been used with good results
for drugs that are poorly water soluble,
and can be optimized for a specific drug
by experimenting with different combina-
tions of excipients. Some aspects of the
optimization process are dependent on the
properties of the specific drug, and thus
optimization will vary from drug to drug.
Lipid-based drug delivery systems
Pharmaceutical drug candidates are often
poorly soluble in water. New enabling drug
delivery systems capable of increasing and
consistency of bioavailability are needed to
accommodate these poorly soluble prod-
ucts. A lipid-based drug delivery system
(LbDDS) is one strategy that takes advan-
tage of the fat-soluble properties of the
drug, rather than fighting against them. Lip-
ids are abundant in the human diet, and the
body is naturally well equipped for lipid pro-
cessing and uptake. Dosing with a LbDDS
resembles the intake of a meal, and has
the potential to increase drug absorption
and eliminate food effect on the molecule’s
bioavailability. LbDDS are already the most
prevalent formulation among drug products
approved from the 1980s through 2015.
Lipid- and surfactant-based drug delivery
systems span a range of categories, includ-
ing oil solutions, emulsions, and surfactant
systems. Whereas dissolution is a crucial
process in formulation of a solid dosage
form, LbDDS bypass dissolution in the
gastrointestinal tract. Instead, digestion and
the formation of complex colloidal phases
within the gut are important considerations
in the design of such a system (1).

16 J ANUA RY 2 0 2 0 | P H A R MT E C H SPO N SO R ED CO N TEN T

Boosting
Bioavail-
Softgel
SNEDDS SEDDS
Formulation
ability
Figure 1: Sample optimization experiment.
Self-nano-emulsifying
drug delivery systems
Self-emulsifying drug delivery systems
(SEDDS) have attracted interest in the
pharmaceutical industry because they can
be dosed as pre-concentrates in a capsule
and generate a drug-containing emulsion
with a large surface area upon dispersion in
the gastrointestinal tract. Some researchers
are taking these emulsions into the realm
of nanotechnology by developing self-nano-
emulsifying drug delivery systems (SNEDDS).
SNEDDS spontaneously form droplets be-
tween 20 and 200 nm in size upon dilution in
water with gentle agitation (2, 3).
17 J ANUA RY 2 0 2 0 | P H A R MT E C H SPO N SO R ED CO N TEN T
Optimizing Lipid-Based Drug Delivery
Anette Müllertz, PhD, professor and head
of Bioneer: FARMA at the Department of
Pharmacy at the University of Copenhagen,
has studied the optimization of LbDDS,
with focus on SNEDDS, extensively. “What
we’re interested in is having a homoge-
neous isotropic preconcentrate that when
we put it into an aqueous phase, rather
spontaneously forms a nano-emulsion or
at least some sort of small lipid droplets,”
said Müllertz at the October 2017 BASF
Solubilization symposium, held in Ludwig-
shafen, Germany.
Müllertz explained that the optimization
process begins with choosing the right ex-
Boosting
Bioavail-
Softgel
SNEDDS SEDDS
Optimizing Lipid-Based Drug Delivery
Formulation
ability

Figure 2: Digestion of lipids in the small intestine.

cipients. The basic components of the formu-
lation include an oil, a triglyceride, and a hy-
drophilic surfactant. Common additions then
include a lipophilic surfactant (co-surfactant)
to form the droplet. A co-solvent is typically
also included in the recipe. In one optimiza-
tion experiment (see Figure 1), Müllertz be-
gan with 13 different excipient compositions
generated by a software algorithm (MODDE
10.1.0 by Umetrics). The resulting lipid droplet
size when dispersed in water was measured.
The goal was to produce a nano-emulsion
with droplets of 40 nm in size or less (shown
in blue in Figure 1) (4).
The optimization experiment showed that
the use of medium-chain triglycerides, as
opposed to long-chain triglycerides, pro-
duced droplets in the correct size range. The
optimal formula included Kolliphor® RH 40,
a nonionic castor oil-based hydrophilic sur-
factant, Lipoid S LPC 80 (soybean-derived,
80.8% monoacyl phosphatidylcholine,
13.2% phosphatidylcholine), and ethanol.
According to Müllertz, the ultimate choice
of SNEDDS formulation for a drug will de-
pend on the solubility of the drug and there-
by the achievable drug load in the droplet.

18 J ANUA RY 2 0 2 0 | P H A R MT E C H SPO N SO R ED CO N TEN T

Boosting
Bioavail-
Softgel
SNEDDS SEDDS
Optimizing Lipid-Based Drug Delivery
Formulation
ability

Figure 3: Formation of multilaminar vesicles during digestion of a

SNEDDS.

The role of digestion processes that are continuing in the small

Gastric lipase is secreted in the stomach,
and is accounting for 10–20% of the lipid
digestion. It digests a single triacylglyc-
eride molecule into one free fatty acid
molecule and a diacylglyceride molecule.
These molecules facilitate emulsification
of lipids before they enter the duodenum.
Thus, any LbDDS will not only undergo
solution-based dispersion in the stomach,
as observed in in vitro dissolution models,
but will also be subjected to digestion
intestine, where the majority of ingested
triglycerides are digested by pancreatic
lipase. Other pancreatic enzymes also
contribute to the digestion of lipids in the
small intestine. In the process, drug con-
tained within the oil droplet is transferred
into vesicles and mixed micelles (see
Figure 2).
There is an assumption that the diges-
tive process will boost the absorption and
bioavailability of a drug in a LbDDS. Little

19 J ANUA RY 2 0 2 0 | P H A R MT E C H SPO N SO R ED CO N TEN T

Boosting
Bioavail-
Softgel
SNEDDS SEDDS
Formulation
ability
Figure 4: In vivo performance in beagle dogs SNEDDS with halofantrine.
experimental evidence has thus far been
gathered about the impact of gastric and
intestinal digestion on the bioavailability of
those drugs, however, Müllertz explained.
In one experiment to study the different
colloidal phases formed during digestion of
a LBDDS, small-angle X-ray scattering (SA
XS) and wide-angle X-ray scattering (WA XS)
were used to visualize the colloidal struc-
tures using a model drug, fenofibrate, and
to evaluate the extent of precipitation of the
drug during digestion (5).
20 J ANUA RY 2 0 2 0 | P H A R MT E C H SPO N SO R ED CO N TEN T
Optimizing Lipid-Based Drug Delivery
Figure 3 shows the formation of mul-
tilaminar vesicles (i.e., vesicles nested
inside each other) during the digestion of a
SNEDDS containing fenofibrate in another
study. Before digestive enzymes are added,
single emulsion droplets are present. After
one minute of digestion, multilaminar
vesicles can be seen as rings within rings in
the x-ray images. The addition of monoacyl
phosphatidylcholine was found to inhibit the
formation of these multilaminar vesicles (6).
The researchers were also able to measure
Boosting
Bioavail-
Softgel
SNEDDS SEDDS
Formulation
ability
Figure 5: Mechanism of absorption from SNEDDS.
precipitation of the model drug in different
formulations in vitro.
The role of solubility
Another prevailing assumption in the
design of LbDDS, including SNEDDS,
is that the drug must be in solution in
the SNEDDS upon dosing in order to be
absorbed. Drug solubility in SNEDDS
is often low, and traditionally, drug is
loaded into SNEDDS at a concentration
of 70–80% of the saturation solubility to
21 J ANUA RY 2 0 2 0 | P H A R MT E C H SPO N SO R ED CO N TEN T
Optimizing Lipid-Based Drug Delivery
reduce the risk of precipitation. Müllertz
and coworkers have investigated the use
of supersaturation in SNEDDS design as a
way of increasing the drug load (9).
Another traditional concern is the pre-
cipitation of drugs in SNEDDS during in
vitro digestion. Studies have shown that
some poorly water-soluble drugs such as
cinnarizine, simvastatin, and halofantrine
precipitated in an amorphous form during
in vitro lipolysis, which results in a faster
dissolution rate compared to the crystal-
line forms of the drugs. This finding calls
Boosting
Bioavail-
Softgel
SNEDDS SEDDS
Optimizing Lipid-Based Drug Delivery
Formulation
ability

into question conventional thinking about “This outcome suggests

the problems of API precipitation (8).
Using halofantrine as a model drug, that the digestion of lipids
Müllertz and collaborators assessed the is not as important in
distribution of the drug between solubilized
and precipitated states after in vitro diges- the absorption of poorly
tion using supersaturated SNEDDS (super- soluble drugs dosed in
SNEDDS) preconcentrate, and subsequently
evaluated the bioavailability in vivo in beagle SNEDDS as previously
dogs. The in vivo study showed that absorp- believed.”
tion of halofantrine was not affected by
in vitro precipitation, and the researchers
the digestion of lipids is not as important
conclude that amorphous precipitation
in the absorption of poorly soluble drugs
may actually enhance absorption. “If a drug
dosed in SNEDDS as previously believed.
precipitates in amorphous [form], you can
The mechanism of absorption may involve
imagine that once it gets into another envi-
partitioning of the drug directly from the
ronment in the intestine, it will very quickly
nanoemulsion droplet to bile salt micelles
go into solution and then it can easily be
(see Figure 5).
absorbed,” said Müllertz.
“We can obviously further question
whether lipolysis is actually needed for drug
absorption, which is sort of an interesting
Revisiting digestion aspect, because that’s going to change the
Challenges to conventional wisdom on the current paradigms if that’s actually true,”
effect of drug precipitation raise additional said Müllertz.
questions about whether assumptions
about the role of digestion are valid. Mül-
lertz investigated the effect of digestion on The chasing principle
drug absorption in rats using tetrahydrolip-
One further strategy that can be employed
statin, an inhibitor of pancreatic lipase, gas-
with SNEDDS is what Müllertz describes as
tric lipase, and carboxyl ester lipase, with
“the Chasing Principle.” That means dosing
the model drug halofantrine (10).
with a drug in suspension and dosing after-
The study validated the super-SNEDDS ef-
ward with a drug-free SNEDDS, or using it
fect that Müllertz had observed in beagles.
as a “chaser.” A study (11) using a cinnari-
In addition, it revealed an extended absorp-
zine aqueous suspension with a SNEDDS
tion phase when digestive enzymes were
chaser compared with traditional SNEDDS
inhibited with tetrahydrolipstatin, but that
dosing strategies showed that in vivo, the
extended absorption phase did not influ-
chasing principle produced the same results
ence the overall bioavailability of the drug
as if the drug was in solution.
(see Figure 4). This outcome suggests that

22 J ANUA RY 2 0 2 0 | P H A R MT E C H SPO N SO R ED CO N TEN T

Boosting
Bioavail-
Softgel
SNEDDS SEDDS
Formulation
ability
That opens up options for dosing, Müllertz
said: “It means that you can basically dose
a tablet together with a lipid dose and then
the drug goes in solution in the stomach.”
Conclusions
SNEDDS have shown to be a useful ap-
proach for dosing of poorly soluble drugs.
One traditional limitation in SNEDDS appli-
cation is the relatively low solubility of many
drugs in SNEDDS. Müllertz and collabora-
tors have shown that this can be overcome
if the drug can be dosed in a supersaturated
state in the SNEDDS (super-SNEDDS).
Super-SNEDDS is a viable option for driv-
ing drug absorption, but do run the risk
that the drug could precipitate in crystalline
form, rather than the more desirable amor-
phous form. A drug that precipitates in crys-
talline form would be subject to the same
need for dissolution as a tablet drug—a goal
that is challenging with poorly soluble API.
Another limitation in the use of SNEDDS
is the assumption that the drug must stay
in solution upon dissolution or digestion in
order to be available absorbed, and that di-
gestion plays a critical role in drug absorp-
tion. Müllertz presented research showing
drug precipitates in amorphous form can
be readily absorbed, and that drug absorp-
tion possibly can occur directly from lipid
particles, without digestion as a mediating
process.
23 J ANUA RY 2 0 2 0 | P H A R MT E C H SPO N SO R ED CO N TEN T
Optimizing Lipid-Based Drug Delivery
The properties of different APIs can
vary dramatically, affecting their suitability
for SNEDDS. The Chasing Principle is a
strategy that may work well for some
drugs. Predictive and in vitro methods are
needed to further improve understanding
of the absorption mechanisms for drug de-
livery systems.
References
(1) A. Müllertz, A. Ogbonna, S. Ren, and T. Rades, J.
Pharm. Pharmacol. 62 (11), 1622–1636 (2010).
(2) A. Date and M. Nagarsenker, Int. J. Pharmaceutics
329 (1–2), 166–172 (2007).
(3) A.A-W. Shahba, K. Mohsin, and F.K. Alanazi, AAPS
PharmSciTechnol. 13 (3), 967–977 (2012).
(4) T. Tran, X. Xi, T. Rades, and A. Müllertz, Int. J. Pharma-
ceutics 502 (1–2), 151–160 (2016).
(5) T. Tran, S. D.v.s. Scheyla, H.A. Siqueira, A. Müllertz,
and T. Rades. J. Controlled Release 255, 45–53 (2017).
(6) T. Tran, D.V.S. Scheyla, H.A. Siqueira, H. Amenitsch,
Th. Rades, and A. Müllertz. Eur. J. Pharm. Sci. 108,
62–70(2017).
(7) A.T. Larsen, et al., Eur. J. Pharm. Sci. 48 (1–2), 339–
350 (2013).
(8) N. Thomas, et al., AAPS J. 16 (3), 539–549 (2014).
(9) N. Thomas, et al., J. Controlled Release 160 (1), 25–32
(2012).
(10) M.H. Michaelsen, et al., AAPS J. 18 (1), 180–186
(2015).
(11) S. D.v.s. Siqueira, et al., AAPS J.19 (2), 587–594
(2017).
Boosting
Softgel
Bioavail-
ability
Formulation
SNEDDS SEDDS
Introduction
The modern solubilization challenge is
that APIs that cannot be dissolved, cannot
be absorbed, and ultimately cannot treat
patients. Poorly water-soluble drugs are
typically divided into two macro catego-
ries, “brick dust” molecules, exhibiting a
strong crystal lattice structure or “grease
ball” molecules, exhibiting notably low
melting point but very high hydrophobicity.
Lipid-based drug delivery systems, specifi-
cally those made from self-emulsifying
drug delivery systems (SEDDS) have been
shown to be an excellent formulation for
the improvement of bioavailability of poor-
ly water-soluble drugs in both categories.
However, the formulation of these
systems is traditionally difficult due to
the temperature and ingredient sensi-
tivity of the formulations. In order to
identify robust, “off-the-shelf“ proven
SEDDS formulations, an advanced, High
Throughput Robotic system was utilized
to systematically identify SEDDS for-
mulations in order to create a tertiary
24 J ANUA RY 2 0 2 0 | P H A R MT E C H SPO N SO R ED CO N TEN T
High Throughput Screening for
the Discovery of SEDDS
S P ONS OR E D C ONTE NT
High Throughput
Screening for the
Discovery of Self-
Emulsifying Drug
Delivery Systems
(SEDDS)
Frank Romanski and Ronny Hoffmann
phase diagram for a series of com-
monly used pharmaceutical ingredients.
These were then subsequently tested
for dispersibility and feasibility for use
in soft and hard capsule systems.
Methods
A BASF-designed high throughput
robotic system (HTS) was utilized to
screen and identify tertiary phase
diagrams capable of forming stable
microemulsions. Samples were heated,
dispersed, mixed, and analyzed in a
precisely defined order and time. The
resulting formulations were tested
for viscosity, turbidity, conductivity,
and image analysis to determine true
microemulsion regions, defined as low
viscosity (≤700 cPs), isotropic and opti-
cally clear systems. An image of the
setup is shown in Figure 1.
Ingredients were dosed by combining
surfactant, cosurfactant, pH 7.0 buff-
ered water, and oil phases. Homogeni-
Boosting High Throughput Screening for
Softgel
Bioavail-
ability
Formulation
SNEDDS SEDDS the Discovery of SEDDS

RH 40 (RH 40) was used as the primary

Figure 1: BASF designed high
surfactant, while co-surfactant Glyceryl
throughput robotic system (HTS).
Monooleate (GMO) was used to modu-
late the HLB value. Each HLB value
served as a surfactant combination
“block.”
A total of 180 experiments were com-
pleted to cover the three-dimensional
design space; these can be exemplified
by Figure 2 and Figure 3:

Figure 2: Surfactant concentra-

tion and composition experimental
design plot.
zation, utilizing 1 min shaking was used
as a first pass mechanism for mixing
the systems. If the systems remained
inhomogeneous after 1 minute, a
20-second homogenization was utilized
using the Ultra Turrax T-25 homogenizer
at 60°C; samples were rested then for
24 hours and tested. Homogeneous
samples were noted, while clearly phas-
eseparate systems were also noted.
Aqueous phases were made with Figure 3: Surfactant concentration
purified water, oil phases with medium and oil phase fraction experimen-
chain triglycerides (Kollisolv® MCT 70), tal design plot.
primary surfactant was Kolliphor ® RH
40, with smaller levels of cosurfactant
glyceryl monooleate. These target
excipients were previously identified
using high throughput chemistry. Five
ratios of surfactant and co-surfactant
were explored to identify the most ef-
fective ratio. These ratios were isolated
by the desired HLB value. Kolliphor ®

25 J ANUA RY 2 0 2 0 | P H A R MT E C H SPO N SO R ED CO N TEN T

Boosting High Throughput Screening for
Softgel
Bioavail-
ability
Formulation
SNEDDS SEDDS the Discovery of SEDDS

Phase Volume ratio is the ratio of oil
to the formulation, defined as:
Surfactant mixtures blocks were de-
fined by Table 1:
Table 1: Surfactant mixtures uti-
lized for optimal microemulsion
discovery.
Samples that were demonstrably unstable
are labeled red.
Microemulsion formulations were dis-
persed in either 0.08 M HCl buffer or 6.8
pH Phosphate buffer to test for dispers-
ibility and use as a SEDDS formulation.
Droplet sizes were tested using dynamic
light scattering (Malvern Zetasizer). Drug
concentrations were determined via HPLC
using two model compounds, danazol and
nifedipin.
Results and discussion
True microemulsions are formed when ex-
act concentrations of water, oil, surfactant
and co-surfactant are utilized within a giv-
en temperature range. These stable zones
are typically described utilizing a classic
“Fish Tail” diagram, shown in Figure 4.

Figure 4: Fish tail diagram for mi-

croemulsion phases space.

All results were then subsequently plot-

ted on tertiary phase diagrams. Success-
ful microemulsion formulations, noted as
isotropic with optical clarity, single phase,
and low viscosity, were noted and sub-
sequently evaluated for consistency and
stability; these are labeled green in the
subsequent charts. Samples that retained
stability after homogenization are labeled
as blue, and while stable are not consid- On the x-axis, the surfactant volume is
ered true bi-continuous microemulsions. shown, as with very high present surface
area in a bi-continuous system, comple-

26 J ANUA RY 2 0 2 0 | P H A R MT E C H SPO N SO R ED CO N TEN T

Boosting High Throughput Screening for
Softgel
Bioavail-
ability
Formulation
SNEDDS SEDDS the Discovery of SEDDS

mentary high levels of surfactants are ing a surfactant ratio of 85:15, RH 40 to

required. The y-axis can be plotted either GMO, a theoretical HLB blend of 13.3,
as temperature or as the surfactant mix- is shown in Figure 6.
ture; in this work, temperature remained
constant and the surfactant mixture was
varied. Winsor Type IV formulations are Figure 6: Block II tertiary phase
the target, where appropriate quantities of diagram – surfactant ratio RH
surfactant, oil, and aqueous phase form a 40:GMO as 85:15.
bi-continuous, isotropic, and low-viscosity
microemulsion. Once formed, they will
remain thermodynamically stable unlike
traditional O/W or W/O emulsions that coal-
lesse over time. As described, five tertiary
systems were constructed by varying con-
centrations of oil, water, surfactant, as well
as surfactant blend. Blend concentrations
are depicted in the five diagram blocks, the In Block II, it is clear that the upper
first, which did not contain a co-surfactant right region with 10% water and varied
is shown in Figure 5. ratios of oil and surfactant was able to
successfully produce microemulsions.
These, labeled in green, vary by a sur-
Figure 5: Block I tertiary phase
factant ratio of 40% to 80% w/w. By
diagram – surfactant ratio RH
40:GMO as 100:0. increasing the level of co-surfactant to
75:25, reducing the HLB to 12.2, the
next diagram was produced as Block III,
shown in Figure 7.

Figure 7: Block III tertiary phase

diagram – surfactant ratio RH
40:GMO as 75:25.
The use of a co-surfactant was clearly
needed in order to create true micro-
emulsions. It is important to note that
the region in the upper right corner,
where large concentrations of Oil and
Surfactant are present with 10% water
were stable posthomogenization. Utiliz-

27 J ANUA RY 2 0 2 0 | P H A R MT E C H SPO N SO R ED CO N TEN T

Boosting
Softgel
Bioavail-
ability
Formulation
SNEDDS SEDDS
What is clear from Block III, as well as
Blocks IV and V, which are not shown
for brevity is that higher levels of co-
surfactant were not necessary and also
reduced the microemulsion region on
the diagram. In Block III, only two points
labeled F and G were microemulsions,
while between them clear phase sepa-
ration was evident. Consequently, only
Block II, with the 85:15 surfactant ratio
was utilized for subsequent testing.
It is important to note the droplet
distribution once in contact with the
body or aqueous testing media. True
microemulsions once in contact with
extraneous water will shift to an O/W
emulsion with fine oil droplets rapidly
dispersed. An example of the previ-
ously depicted formulations is shown in
Figure 8, where oil droplet particle size
distribution was tested by dynamic light
scattering in acidic stomach conditions
(HCl buffer, 0.08 M) and intestinal condi-
tions (pH 6.8 Phosphate Buffer).
Figure 8: D50 oil droplet size post
dispersion of microemulsions in
acidic or phosphate
buffer.
28 J ANUA RY 2 0 2 0 | P H A R MT E C H SPO N SO R ED CO N TEN T
High Throughput Screening for
the Discovery of SEDDS
The droplet sizes for D50 are notably
small, with ranges from 15 to 35 nm
at the D50 level. It is important to note
that the distributions were relatively
monodisperse, with D10 ranging from
10 to 20 nm and D90 from 20 to 60
nm, respective to A through D. What
is also clear is that there was no effect
of pH, which is due to the use of non-
ionic surfactants. In a real formulation,
these nanoscale oil droplets would be
rapidly digested allowing for absorption
of the poorly soluble API. In this range,
it is clear that higher oil concentrations
result in larger droplet sizes, where
formulation D with 40% MCT oil has
roughly double the median droplet sizes
as formulation A with only 10% oil. On
the other hand, higher surfactant con-
centrations allow for higher drug load-
ing; this is shown in Figure 9, where
common poorly water soluble drugs da-
nazol and nifedipin are shown at stable
drug loading concentrations.
Figure 9: Drug capacity as a func-
tion of formulation for representa-
tive poorly soluble drugs
danazol and nifedipin.
Boosting
Softgel
Bioavail-
ability
Formulation
SNEDDS SEDDS
Figure 10: Example oil droplet size
distribution for formulations satu-
rated with representative poorly
soluble drug nifedipin.
It is important to note that drug load-
ing did not have a significant effect
on droplet size distribution, one such
example is shown in Figure 10 for nife-
dipin under stomach conditions.
Conclusion
• SEDDS formulations start with a true
microemulsion, which is challenging
and manually cumbersome to identify;
this challenge was addressed utilizing a
high throughput robotic system in order
to effectively identify regions of stable
microemulsions to be used in SEDDS.
29 J ANUA RY 2 0 2 0 | P H A R MT E C H SPO N SO R ED CO N TEN T
High Throughput Screening for
the Discovery
• It is important to utilize a cosurfactant
and subsequently at the right ratio in or-
der to create the most robust space for
microemulsions in an oil-water-surfac-
tant system.
• Aqueous phases are included to account
for moisture absorption and retention in
encapsulated formulations
• Produced microemulsions self-emulsify
as SEDDS systems once dispersed into
aqueous or biorelevant media into nano-
meter scale oil droplets.
• Oil droplet size was not affected by
system pH due to the exclusive use of
non-ionic surfactants.
• Larger droplets are evident with larger
oil ratios in the SEDDS formulation,
while higher drug loading is possible
with larger surfactant ratios.
• This work will be used to further study
the produced SEDDS formulations in
mechanistic testing.
Frank Romanski, PhD, is a global technical
marketing manager at BASF, and Ronny
Hoffmann is a senior scientist at BASF.