The numbers of poorly water-soluble drug candidateshave increased significantly during the past decade.The ability to enhance the apparent solubility ofthese drugs and to stabilize their supersaturated so-lutions is critically important and is needed to ensurethe success in formulating therapeutically effectivedosage forms.1–3 Many formulation techniques suchas solid lipid nanoparticles, self-emulsifying drug de-livery systems, cocrystal formation, and so forth havebeen used to increase the solubility of poorly solu-ble drugs.4–6 Relative to these techniques, delivering
drugs in their amorphous form have gained consider-able attention because of its ability to enhance solu-bility of large number of poorly soluble drug moleculeswith variable physiochemical properties.7–9 However,amorphous drug recrystallization in solid state andprecipitation upon solubilization are common chal-lenges associated with this approach.1,2,9,10 Under-standing these processes in both solution and solidstate is vital for the overall performance of amor-phous dosage forms during drug development stages.Upon dissolving amorphous forms of many drugs, aninitial high supersaturation level is created and ifthis level of supersaturation is maintained, it will al-low for a significant increase in absorption and thushelp achieving sufficient bioavailability of the drug.This generation and maintenance of supersaturationcan be accomplished by various polymers that act
JOURNAL OF PHARMACEUTICAL SCIENCES 1
2 CHAUHAN, HUI-GU, AND ATEF
as precipitation inhibitors from solutions and inter-fere with the drug nucleation and/or crystal growthin solid state.2,3 Effective uses of polymers as precip-itate inhibitor and solid-state stabilizers can be ac-complished by solid dispersion technique, in whichthe drug molecules are embedded in polymeric ma-trix.
In solution state, factors such as pH, viscosity, anddrug–polymer molecular interaction can significantlyaffect polymer precipitation inhibition efficiency ofpolymers, whereas factors such as the drug recrystal-lization tendency, polymer’s molecular mobility, andthe drug–polymer molecular interactions are foundto play a significant role in the recrystallization in-hibition in solid state.8–11 The limiting factor in thebroad utilization of the solid dispersions techniqueis the lack in complete understanding of the drugprecipitation mechanism and the influence of spe-cific polymers on the precipitation kinetics. Although,many attempts have been made to investigate theprecipitation mechanism in solutions,12,13 the com-prehensive understanding of the effect of polymersand different precipitation condition on the final pre-cipitate morphology and crystal form are yet to befully explored.14,15 Also, thorough understanding ofthe drug–polymer molecular interactions and its cor-relation to the solution and solid states stabilizationwould be a huge step forward in the rational screeningof polymers for solid dispersion preparations.
Dipyridamole (DPD), an antiplatelet agent and va-sodilator, was used as a model drug in this study(Fig. 1). It is a class II, poorly soluble, and weakly basiccompound. The drug precipitates after reaching smallintestine because of increased pH resulting in the de-crease of overall absorption and bioavailability.16–20
The objectives of the study are to understand DPDprecipitation mechanism, to study the effects of poly-mers on the crystal form, and morphology as a meanto understand the precipitation interference mecha-nism, then correlating the effect of polymers on drugsupersaturation in solution to solid-state recrystal-lization inhibition.
MATERIALS AND METHODS
Dipyridamole was purchased from MP Biochemi-cals (Solon, Ohio). PVP K90 was purchased fromSpectrum (New Brunswick, New Jersey). Hydrox-ypropylmethylcellulose (HPMC) was purchased fromDow Company (Midland, Michigan). Eudragits (E100,S100, and L100) were purchased from Degussa(Parsippany, New Jersey); and PEG 8000 was pur-chased from Sigma (St. Louis, Missouri). Acetonitrile(ACN), ethanol, methanol, methylene chloride, andhydrochloric acid of analytical or high performanceliquid chromatography (HPLC) grades were used. Allmaterials were used as received.
Figure 1. Structure of dipyridamole.
Solubility Determination
An excess amount of DPD was added to the precipi-tating medium and stirred for 24 and 48 h at 25◦C.The solutions were filtered through 0.45 :m mem-brane filters, then the concentration was determinedusing Agilent(Santa Clara, CA) 1200 series HPLC,with Chemstation software. The mobile phase (A) con-sisting 95% water, 4.9% ACN, and 0.1% trifluoroaceticacid (TFA) and mobile phase (B) consisting 95% ACN,4.9% water, and 0.1% TFA were used. The gradientwas changed from 99% mobile phase A to 99% mobilephase B linearly in 15 min using a flow rate of 1 mL/min. The injection volume was 10 :L, and a wave-length (λ) of 280 nm was used. C-18, 3.5 :m, 4.6 ×50 mm2 column (Waters, Milford, Massachusetts) wasused. All samples were run in triplicates.
Precipitation Studies
In precipitation studies, 5 mL of DPD solution (2 mg/mL in absolute ethanol) was added to 200 mL of phos-phate buffer (pH 6.8) at a rate of 1 mL/min using asyringe pump (Harvard 33’ Syringe Pump, Holliston,Massachusetts). Real-time absorbance and turbidityof DPD in the solution phase were determined at awavelength (λ) of 280 and 500 nm, respectively, usingCary 50 UV–visible continuous spectrophotometer(Varian Analytical Instrument, Palo Alto, California)equipped with fiber optic probe. For drug precipita-tion inhibition studies, PVP K90, HPMC, EudragitE100, Eudragit S100, Eudragit L100, and PEG 8000were used at 1:1 and 5:1 drug–monomer ratio. Theinitial and final pH and viscosity were measured us-ing Orion 3 star pH meter (Thermo scientific, Bev-erly, Massachusetts) and Brookfield DV 2 Viscometer(Middleboro, Massachusetts), respectively.
Drug supersaturation experiments are carried outin a DPD concentration range of 1–4 mg/mL. Therelative supersaturation was calculated based on the
JOURNAL OF PHARMACEUTICAL SCIENCES DOI 10.1002/jps
THE BEHAVIOR OF POLYMERS IN AMORPHOUS STABILIZATION ABILITY 3
Figure 2. Determination of (a) precipitation initiation time from slope change of absorbanceor turbidity curves with time, and (b) precipitation rates derived from concentration versustime. Concentration versus time curves are derived from dipyridamole UV–visible continuousabsorbance experiments (absorbance vs. time).
following equation:
S =(
C − C∗C∗
)(1)
where S is relative supersaturation, C and C∗ arethe concentration and solubility of DPD, respectively.The seeding experiments were carried out by adding0.5 mg of gamma crystalline DPD to 200 mL of phos-phate buffer (pH 6.8) during the precipitation studies.
The precipitated solution was filtered through0.45 :m filters and dried overnight in vacuum oven.The precipitate was characterized by X-ray powderdiffraction (XRPD), modulated differential scanningcalorimetry (MDSC), infrared (IR) spectroscopy, andscanning electron microscopy (SEM).
Calculation of the Precipitation Initiation Timeand the Precipitation Rate of DPD
The DPD precipitation initiation time in the absenceand presence of polymers were calculated by deter-mining the changes in absorbance or turbidity curves(Fig. 2). During precipitation the DPD concentrationdecreased, whereas the precipitated DPD increased,both resulted in observable changes in their respec-tive curves. A correction was applied to count for thefirst 5 min used in adding the DPD solution to theprecipitation medium. The initiation time is definedas the first detectable change in the slope of the ab-sorbance curves. The precipitation rate, defined as thenumber of molecules of DPD precipitating per unitvolume, per unit time, was calculated at different su-persaturations in the presence and absence of poly-mers (Fig. 2).
DOI 10.1002/jps JOURNAL OF PHARMACEUTICAL SCIENCES
4 CHAUHAN, HUI-GU, AND ATEF
Solid-State Studies
Solid dispersions of DPD with various polymers(5:1, 3:1, and 1:1, w/w, drug–polymer) were pre-pared by solvent evaporation technique using a Ro-tavapor (Buchi rotavapor R 200 series; New Castle,Delaware). Methanol and the mixture of methanoland methylene chloride were used as solvents. Thelatter was utilized to improve the solubility of fewmethanol insoluble polymers. The collected disper-sions were sieved and dried in vacuum oven for 24 h.The physical mixtures and amorphous physical mix-tures of DPD were prepared by geometric mixing ina mortar and a pestle. Physical mixtures consist ofcrystalline drug polymer mixtures, whereas amor-phous physical mixtures were prepared using amor-phous form of drug mixed with polymers. The disper-sions and the physical mixtures were characterizedby XRPD, MDSC, IR, and Raman. Solid-state stabil-ity of DPD dispersions was carried out at 25◦C for1 month.
X-Ray Powder Diffraction
Samples were analyzed by XRPD, Bruker AXS-XRD(Billerica, Massachusetts), using Cu K" radiation.The XRPD patterns were collected in the angularrange of 1 < 2θ < 40◦ in step scan mode and ana-lyzed by EVA software (Bruker).
Modulated Differential Scanning Calorimetry
The thermal analysis was carried out using MDSC (Q2000 series, TA Instruments, New Castle, Delaware)equipped with a liquid nitrogen cooling assembly.Samples of 5–10 mg were prepared in sealed pans.The samples were equilibrated to 10◦C, modulated at±0.32 s every 60 s, kept isothermal for 5 min, andramp at 2◦C/min to 180◦C. The data were analyzedby TA Universal Analysis software (TA Instruments).
IR Spectroscopy
Attenuated total reflectance mode was used to obtainIR spectra using Bruker Vertex series 80 V IR spec-trometer. Fifty scans were collected for each sampleover a wave number region of 400–4000 cm−1. IR spec-troscopy was used for the characterization of the pre-cipitate and to explore the drug–polymer interactions.
Raman Spectroscopy
Raman spectra were obtained using Bruker Vertex se-ries 80 V Raman spectrometer; and data processingwas performed using OPUS software (Bruker, Biller-ica, Massachusetts). Data acquisition was carried outusing an exposure time of 10 s for 10 accumulationsand a laser power setting of 400 mW. Raman spec-troscopy was used for the characterization of precipi-tate and to explore the drug–polymer interactions in
various mixtures and dispersions along with IR spec-troscopy.
Scanning Electron Microscopy
Scanning electron microscopy images were acquiredusing scanning electron microscope Magellon XHRSeries (Hillsboro, Oregon) operating between 5 and24 kV. The specimens were mounted on a metal stub,with double-side adhesive tape and coated under vac-uum with gold in an argon atmosphere before obser-vation. SEM was used to study the morphology of aprecipitate.
RESULTS
DPD Precipitation in the Absence of Polymers
The solubility of DPD was found to be 3.1 :g/mL inthe precipitating medium. No difference in solubilitywas observed at these time intervals. Using Eq. 1, thecalculated supersaturation levels ranged from 6.8 to30.5 for different amount of DPD (1–4 mg/mL) addedduring precipitation experiments. At supersaturationbelow 4.5, no precipitation was observed for 300 min,whereas at a supersaturation greater than 30.5, theprecipitate was immediately formed during the ad-dition of DPD solution. No significant difference wasobserved in absorbance and turbidity changes in thesolution phase at wavelength (λ) of 280 and 500 nm,respectively. It was confirmed that the concentrationdrop in DPD at 280 nm resulted in turbidity at 500nm and both absorbance and turbidity curves could beeffectively used for the determination of precipitationinitiation time and rates. The concentration change ofDPD obtained from absorbance curve at 280 nm wasused in calculating the initiation time and precipita-tion rate. A linear inverse proportionality was foundbetween the precipitation initiation time and super-saturation (Fig. 3). Similarly, the DPD average precip-itation rates also increased with increase in supersat-uration. These rates may correspond to simultaneouscrystal nucleation and growth rates as it is very hardto distinguish between these two events.1,12,13 It wasalso observed that the precipitation rate, although un-expected, significantly decreased with seeding, whichis contradictory to the precipitation in the absences ofseeds, at the same supersaturation level. The resultswill be explored more in future research. Cuboidalcrystal habit was observed at all supersaturationswithout seeds. Seeding resulted in an increased ir-regularity of the precipitated DPD (Fig. 4).
Polymer’s Effects on DPD Precipitation
The Effect of Polymers on the Precipitation InitiationTime
The initial screening of polymers at 1:1 (drug—monomer) ratio in solution showed that Eudragit
JOURNAL OF PHARMACEUTICAL SCIENCES DOI 10.1002/jps
THE BEHAVIOR OF POLYMERS IN AMORPHOUS STABILIZATION ABILITY 5
Figure 3. Relative supersaturation versus initiation time(initiation time of DPD precipitation linearly decreases withincrease in supersaturation).
E100, HPMC, and Eudragit S100 increased the pre-cipitation initiation time to greater than 600, 380 ±25.4, and 65 ± 8.4 min, respectively, compared with16 ± 10.2 min in the absence of the polymers. No pre-cipitate was observed with Eudragit E100 during thetime duration of experiment, that is, 600 min. Otherpolymer’s PVP K90, Eudragit L100, and PEG 8000 didnot show any significant increase in the precipitationinitiation time at 1:1 (drug–monomer) ratio (Fig. 5).Similar results were observed at 5:1 (drug–monomer)ratio for all the polymers. Neither the pH nor theviscosity changed throughout the experiment at thestudied drug–to–monomer ratios. Thus, the rank or-der of the polymers precipitation inhibition efficiencybased on initiation time was found to be Eudragit
At 1:1 and 5:1 (drug–monomer) ratios, HPMC, Eu-dragit S100, and PVP K90 showed a significant de-crease in the DPD precipitation rates. At the sameconcentration, Eudragit L100 demonstrated a slightdecrease in the precipitation rates, whereas PEG8000 showed no effect on the DPD precipitation rates.Thus, the rank order of the polymers’ precipitationinhibition efficiency based on precipitation rates wasfound to be HPMC > Eudragit S100 > PVP K90 > Eu-dragit L100 > PEG 8000. Precipitation rates in thepresence of Eudragit E100 at these concentrationswere not considered, as it did not precipitate out dur-ing the experiments.
The Effect of Polymers on the Crystal Formand the Morphology of the Precipitate
Polymers (Eudragit E100, HPMC, and EudragitS100) with significant effect on the precipitation ini-tiation time resulted in crystal form and morphologychanges of DPD precipitate. Changes in the IR spectraof DPD precipitates in the presence of these polymersat 1500–1550, 1400–1480, and 1350 cm−1 were ob-served. These changes correspond to (C N) stretch,C O stretching plus OH deformation combinationband, and CH deformation vibration bands. Further,MDSC thermograms showed two melting endothermsin the presence of these polymers at 120–150◦Cand 160–165◦C (Fig. 6). These polymers resulted in
Figure 4. SEM of precipitated DPD at various supersaturation levels without seeding (toppictures) and with seeding (bottom figures) at the same supersaturation level. Without seed-ing, there are no notable changes in morphology at different supersaturation levels, but moreirregularities were observed with seeding.
DOI 10.1002/jps JOURNAL OF PHARMACEUTICAL SCIENCES
6 CHAUHAN, HUI-GU, AND ATEF
Figure 5. Polymers effect on DPD precipitation initiation time (symbol “∗” represents signif-icant difference between initiation time in the presence and absence of polymers, calculatedusing one-way analysis of variance; p < 0.001).
Figure 6. MDSC of DPD precipitated in the presence of various polymers: (a) Eudragit E100,(b)HPMC, (c) Eudragit S100, (d) PVP K90, (e) Eudragit L100, and (f) PEG 8000. Additionalendotherm observed in the presence of Eudragit E100, HPMC, and Eudragit S100 is highlightedby the dotted circle in the figure. This additional endotherm corresponds to the conversion ofDPD polymorphic form II to I.
JOURNAL OF PHARMACEUTICAL SCIENCES DOI 10.1002/jps
THE BEHAVIOR OF POLYMERS IN AMORPHOUS STABILIZATION ABILITY 7
Figure 7. SEM of the DPD precipitate in the presence of different polymers at 1:1 (drug–monomer) ratio at S = 14.7 (significant change in morphology was observed in the presence ofEudragit E100, HPMC, and Eudragit S100).
morphological changes in the of DPD precipitate fromcuboidal to irregular crystals (Fig. 7).
On the contrary, no change in the crystal form,confirmed by XPRD and IR, or morphology was ob-served with polymers showing no effect on precipi-tation initiation time (PVP K90, Eudragit L100, andPEG 8000) (Figs. 6 and 7). Also, a melting endothermat 160–165◦C similar to pure DPD was observed inthe thermograms of the precipitates collected in thepresence of these polymers. PVP K90, which has asignificant effect on the precipitation rate of DPD butnot the initiation time, showed no effect on the precip-itate crystal form or the morphology. Similar resultswere observed at various relative supersaturations.
Polymer’s Effects on the Solid-State Stability of DPDDispersion
No changes in the XRPD patterns, IR, and Ramanspectra of the physical mixtures were observed, con-firming the absence of interaction between DPD andpolymers. XRPDs of physical mixtures and solid dis-persions were shown in Figure 8. Further, Table1 summarizes the characterization results of amor-phous solid dispersions prepared with various poly-mers. Stable amorphous solid dispersions were ob-tained with Eudragit S100 at 5:1, 3:1, and 1:1 (w/w)(drug–polymer) ratios. PVP K90, HPMC, and Eu-dragit L100 were found to be less efficient andformed amorphous DPD solid dispersions only at 3:1and 1:1 (w/w) polymer concentrations. No conversion
from amorphous to crystalline form was observedfor 1 month at 25◦C in all amorphous dispersions.Solid dispersions of DPD with Eudragit E100 andPEG 8000 were crystallized during the preparationat all polymers concentrations. The rank order of thepolymers based on minimum polymer concentrationneeded to stabilize the amorphous drug was foundto be Eudragit S100 > PVP K90, HPMC, EudragitL100 > Eudragit E100, PEG 8000.
Dipyridamole solid dispersions with Eudragit S100and HPMC showed changes in IR and Raman spec-tra in both 2750–3000 and 1300–1700 cm−1 re-gions (Figs. 9–11). In the 1300–1700 cm−1 region,the 1534 and 1360 cm−1 peaks correspond to C Nring and C N bonds, respectively, whereas peaks at2932 and 2851 cm−1 are assigned to asymmetricaland symmetrical stretch of CH2 group of DPD.21 Onthe basis of changes in the IR spectra, it could bepostulated that carbonyl group (C O) of polymersmight be interacting with the C N or O H bondpresent in DPD through hydrogen bonding. Simi-lar results were obtained by Beten et al.22 whereinthey found a hydrogen-bonding interaction betweenthe carboxylic function of polymer (Eudragit S100)and the nitrogen atom of DPD through NMR. Nochanges are observed in case of solid dispersions pre-pared from the remaining polymers, that is, PVPK90, Eudragit E100, Eudragit L100, and PEG 8000.Table 2 summarizes all the solution and solid-statestudies.
DOI 10.1002/jps JOURNAL OF PHARMACEUTICAL SCIENCES
8 CHAUHAN, HUI-GU, AND ATEF
Figure 8. (a) XRPD pattern of DPD physical mixtures 1:1 (w/w) drug–polymer ratios withPVP K 90, HPMC, Eudragit E 100, Eudragit S 100, and Eudragit L 100 (bottom to top); (b)XRPD pattern of DPD solid dispersions 5:1 (w/w) drug–polymer ratios with pure DPD, PVP K90, HPMC, Eudragit E 100, Eudragit S 100, and Eudragit L 100, PEG 8000 (bottom to top);and (c) XRPD pattern of DPD solid dispersions 1:1 (w/w) drug–polymer ratios with PVP K 90,HPMC, Eudragit E 100, Eudragit S 100, and Eudragit L 100, PEG 8000 (bottom to top).
Table 1. Characterization of DPD Solid Dispersions by XRPDand MDSC
aAll amorphous solid dispersions were found to be stable for 1 monthat 25◦C.
DISCUSSION
Previous studies on DPD during in vitro precipitationtesting discussed the complexity of the phenomenonbecause of the numerous factors involved in the su-persaturation process and could affect the kinetics ofDPD precipitation such as the gastric-emptying rate,the fed or fasted state, and so forth.17,23 In this work,we systematically studied the precipitation of DPD in
solution and found that DPD has a wide metastablezone and low precipitation potential because no pre-cipitate was observed below S 4.5 (supersaturation of4.5) for 300 min. Similar to our results, Gu et al.16 alsofound that DPD has a low precipitation potential com-pared with another poorly soluble basic drugs such ascinnarzine.
A linear decrease in the precipitation initiationtime with supersaturation confirmed the fact thatthe supersaturation is the driving force of DPD pre-cipitation. Increase in supersaturation, as expected,is found to increase the DPD crystal nucleation andcrystal growth rates as seen by DPD initiation timeand precipitation rates. DPD precipitation has pre-viously been successfully simulated by using simpleclassical nucleation theory in computational oral ab-sorption simulation.24 Our study showed that highconcentration of DPD resulted in fast precipitationof drug. Decreasing the supersaturation of DPD con-siderably decreased the precipitation rates. Inter-estingly, supersaturation have no effect on crystalform and morphology of the DPD precipitate unlike
JOURNAL OF PHARMACEUTICAL SCIENCES DOI 10.1002/jps
THE BEHAVIOR OF POLYMERS IN AMORPHOUS STABILIZATION ABILITY 9
Figure 9. A comparison of IR Spectra between crystalline DPD, crystalline DPD–EudragitS100 physical mixture, amorphous DPD–Eudragit S100 physical mixture, and DPD–EudragitS 100 solid dispersions at 3:1 (drug–polymer) ratios in the region of (a) 2750–3000 cm-1 and (b)1800–1300 cm-1.
Figure 10. A comparison of IR Spectra between crystalline DPD, crystalline DPD–HPMCphysical mixture, amorphous DPD–HPMC physical mixture, and DPD–HPMC solid dispersionsat 3:1 (drug–polymer) ratios in the region of (a) 2750–3000 cm-1 and (b) 1600–1300 cm-1.
Figure 11. A comparison of Raman Spectra between (a) crystalline DPD, crystalline DPD–Eudragit S100 physical mixture, amorphous DPD–Eudragit S100 physical mixture, and DPDEudragit–S100 solid dispersion; and (b) crystalline DPD, crystalline DPD–HPMC physicalmixture, amorphous DPD–HPMC physical mixture, and DPD–HPMC solid dispersion at 3:1(drug–polymer) ratios in the region of 2750–3000 cm-1.
DOI 10.1002/jps JOURNAL OF PHARMACEUTICAL SCIENCES
10 CHAUHAN, HUI-GU, AND ATEF
Tab
le2.
Su
mm
ary
ofS
olu
tion
and
Sol
id-S
tate
Stu
dies
Sol
uti
onS
tudi
esS
olid
-Sta
teS
tabi
lity
Pol
ymer
Init
iati
onT
ime
Pre
cipi
tati
onR
ate
Mor
phol
ogy
and
Cry
stal
For
mC
han
geC
omm
ent
Am
orph
ous
Sta
bili
zati
onM
olec
ula
rIn
tera
ctio
nG
lass
Tra
nsi
tion
Tem
pera
ture
(◦C
)C
omm
ent
PV
PK
90X
a�
�b
XS
teri
ch
indr
ance
�X
�(1
72)
Gla
sstr
ansi
tion
HP
MC
��
��
��
�M
olec
ula
rin
tera
ctio
n�
��
(154
)M
olec
ula
rin
tera
ctio
nan
dgl
ass
tran
siti
onte
mpe
ratu
reE
udr
agit
E10
0�
��
X�
Mol
ecu
lar
inte
ract
ion
XX
X(4
5)N
oco
rrel
atio
nE
udr
agit
S10
0�
��
��
Mol
ecu
lar
inte
ract
ion
��
��
(165
)M
olec
ula
rin
tera
ctio
nan
dgl
ass
tran
siti
onte
mpe
ratu
reE
udr
agit
L10
0X
�X
Ste
ric
hin
dran
ce�
X�
(150
)G
lass
tran
siti
onP
EG
8000
cX
XX
No
effe
ctX
X-
No
effe
ct
aX
,pol
ymer
ssh
owin
gn
oef
fect
.b�
,pol
ymer
ssh
owin
gef
fect
.c P
EG
8000
isa
crys
tall
ine
poly
mer
.
many poorly soluble drugs that precipitate inmetastable form at high supersaturation levels suchas indomethacin.25 For DPD, it appears that super-saturatuion increases the precipitation rates but doesnot influence the mechanism of precipitation. Thiscorrelates well with the results obtained by Arnoldet al. wherein no change in morphology was ob-served with different transfer rates (simulated gas-tric medium to simulated intestinal medium) of 4and 9 mL/min using a novel in vitro drug testing in-strument. Using mathematical modeling, same nu-cleation exponent of 5 and growth exponent of 1.5 inboth the transfer rates was seen.26
In the presence of low polymer concentration, somepolymers were reported to have a significant effecton the drug precipitation. For Example, Adhiyamanand Basu27 reported different crystal modificationsof DPD using 2% solutions of various surfactants andpolymers. The dissolution rates of these newly formedcrystals were found to be greater than pure DPD. Itwas postulated that the changes in the DPD crystalform is because of the adsorption of these excipientsmolecules on the DPD crystal surface.27 In our study,we found that the polymer’s effect on DPD precipita-tion is because of drug–polymer interactions becauseno changes were observed in other experimental fac-tors such as pH and viscosity. Further, these polymerswere found to have different precipitation inhibitionmechanisms because polymers exhibited different ef-fects on the nucleation (initiation time) and crystalgrowth (precipitation rates). On the basis of our study,polymers such as Eudragit E100 and HPMC are foundto be highly efficient inhibitors of DPD precipitationand can be utilized in various supersaturated sys-tems for precipitation inhibition. It is valuable torank order the polymers based on their effectivenessin solid and solution stabilization of the amorphousform. Such information is valuable in the rationalselection of polymers for formulation development,particularly in case of poorly soluble drugs. Earlyscreening of these polymers is needed to select thebest polymer for solubility enhancement. Previously,Avdeef et. al.28 used the high throughput screening(HTS) approach for drugs intrinsic solubility screen-ing in the presence of various excipients. They usedeight sparingly soluble compounds (acidic, basic, andneutral compounds including DPD) and found variousexcipients that can be use to enhance solubility. But,the HTS approach does not provided any informationabout the key factors influencing the precipitationmechanism. Two of these key factors are the precip-itated crystal form and morphology. In the presenceof polymers, pharmaceutical compounds often precip-itate into solvate, hydrate, or different polymorphicform. Morphological changes are also common in thepresence of these excipients and polymers. Both crys-tal form and morphology can significantly affect the
JOURNAL OF PHARMACEUTICAL SCIENCES DOI 10.1002/jps
THE BEHAVIOR OF POLYMERS IN AMORPHOUS STABILIZATION ABILITY 11
solubility and dissolution of poorly soluble drugs.27 Inour study, we explored these two parameters as well.Interestingly, all three polymers that delayed the pre-cipitation initiation time also resulted in DPD crystalform and morphological changes. The IR spectra ofDPD precipitate in the presence of Eudragit E100,HPMC, and Eudragit S100 showed spectral changescompared with the precipitated DPD spectrum, inthe absence of polymers, indicating change in theenvironment of the functional groups and thus sug-gesting drug–polymer interactions. Because differentDPD IR peak shifts were observed in the presenceof the above-mentioned polymers, it reflects differ-ent drug–polymer interactions. MDSC also confirmsthis finding as precipitates in the presence of thesepolymers (Eudragit E100, HPMC, and Eudragit S100)showed an additional melting endotherm in between130◦C and 140◦C with variable enthalpy of melting.This additional endotherm could correspond to theconversion of DPD polymorphic form II to I. DPDform II is thermodynamically stable at room tempera-ture and transforms at around 130◦C to form I, whichmelts at 162◦C (Fig. 6).29 Overall, these changes couldbe a result of difference in strength or sites of molec-ular interaction (functional groups of drug and poly-mers involved in interaction).
In our study, polymers such as PVP K90 and Eu-dragit L100 showed no change in the crystal formand crystal morphology, although these polymers arefound to affect precipitation rates. The decrease in theprecipitation rates of DPD molecules by these poly-mers could be explained because of steric hindranceowing to their high molecular weight and long chainlengths. PEG 8000 showed no effect on DPD precipi-tation as well as its precipitate form and morphology.Possible steric hindrance because of the low molecularweight of PEG 8000 could easily be ruled out.
Amorphous stabilization of poorly soluble drugshas previously been correlated to its glass transi-tion temperature and molecular interactions.30–32 Insome cases, molecular interactions, whereas in othercases, increase in the glass transition temperaturewas found to play a prominent role in stabilizingamorphous drugs. Previous studies have shown thestabilization of amorphous drug by various polymersbut the stabilization mechanism was not explored indetails. Because high polymer concentrations wereused in these studies, the stabilization is most likelybe achieved either by reduced molecular mobility re-flected by an increase in glass transition temperatureor because of some molecular interactions.33–36
In our solid-state study, Eudragit S100 was effec-tive at the lowest polymer–drug ratio of 5:1 (w/w)(DPD–Polymer). Other polymers, with similar glasstransition temperature, were not effective in stabi-lizing solid dispersions at similar polymer concentra-tions. This confirms that at this low polymers concen-
tration, the Eudragit S100 stabilization effect is pri-marily because of the drug–polymer interaction (con-firmed by IR and Raman spectra). Solid dispersionswere successfully formed using polymers such as Eu-dragit L100, PVP K90, and HPMC (high glass tran-sition temperature) when the concentrations of poly-mers were increased to 3:1 and 1:1 (w/w). This couldbe attributed to decreased molecular mobility of theseformed amorphous solid dispersions. DPD dispersionswere not formed with Eudragit E100 (low glass tran-sition temperature), although it showed a strong ten-dency toward molecular interaction during our so-lution precipitation studies (Fig. 5). These resultsshowed that both the glass transition and the molec-ular interaction play significant role and are neededfor the stabilizing of DPD amorphous dispersions. Eu-dragit S100 was found to be effective at the lowestdrug–monomer ratio and could be considered as themost effective stabilizing polymer for the DPD disper-sions. Eudragit S100 stabilization efficiency could becorrelated to stabilization through both molecular in-teraction and increased glass transition temperature.The rank order of polymers based on their efficiencyof amorphous stabilization in solid state was found tobe Eudragit S100 followed by HPMC, PVP K90, andEudragit L100. It correlates well with DPD solutionresults except for Eudragit E100. The later showedeffective stabilization of the supersaturated solutionsbut not the solid-state amorphous stabilization. Thislimited effect is possibly because of the role playedby glass transition temperature in DPD solid-stateamorphous stabilization.
CONCLUSIONS
The polymers type and their ability to preventprecipitation are directly related to polymers abil-ity to produce changes in crystal form (conversionto metastable polymorph II) and its morphology(cuboidal to irregular). Drug–polymer interaction isbelieved to play a significant role in the precipitationinhibition of DPD in solution and amorphous stabi-lization in solid state. Glass transition temperaturealso plays an important role in solid-state stabiliza-tion. Overall, solution and solid state results corre-lated well except Eudragit E100.
REFERENCES
1. Brouwers J, Brewster ME, Augustijns P. 2009. Supersaturat-ing drug delivery systems: The answer to solubility-limitedoral bioavailability? J Pharm Sci 98(8):2549–2572.
2. Curatolo W, Nightingale JA, Herbig SM. 2009. Utility of hy-droxypropylmethylcellulose acetate succinate (HPMCAS) forinitiation and maintenance of drug supersaturation in the GImilieu. Pharm Res 26(6):1419–1431.
3. Guzman HR, Tawa M, Zhang Z, Ratanabanangkoon P, Shaw P,Gardner CR, Chen H, Moreau JP, Almarsson O, Remenar JF.
DOI 10.1002/jps JOURNAL OF PHARMACEUTICAL SCIENCES
12 CHAUHAN, HUI-GU, AND ATEF
2007. Combined use of crystalline salt forms and precipitationinhibitors to improve oral absorption of celecoxib from solidoral formulations. J Pharm Sci 96(10):2686–2702.
4. Kawabata Y, Wada K, Nakatani M, Yamada S, Onoue S. 2011.Formulation design for poorly water-soluble drugs based onbiopharmaceutics classification system: Basic approaches andpractical applications. Int J Pharm 420(1):1–10.
5. Kawakami K. 2012. Modification of physicochemical charac-teristics of active pharmaceutical ingredients and applicationof supersaturatable dosage forms for improving bioavailabilityof poorly absorbed drugs. Adv Drug Deliv Rev 64(6):480–495.
6. Shegokar R, Muller RH. 2010. Nanocrystals: Industrially fea-sible multifunctional formulation technology for poorly solubleactives. Int J Pharm 399(1–2):129–139.
7. Bikiaris DN. 2011. Solid dispersions, part I: Recent evolutionsand future opportunities in manufacturing methods for disso-lution rate enhancement of poorly water-soluble drugs. ExpertOpin Drug Deliv 8(11):1501–1519.
8. Serajuddin AT. 1999. Solid dispersion of poorly water-solubledrugs: Early promises, subsequent problems, and recentbreakthroughs. J Pharm Sci 88(10):1058–1066.
9. Yu L. 2001. Amorphous pharmaceutical solids: Prepara-tion, characterization and stabilization. Adv Drug Deliv Rev48(1):27–42.
10. Hancock BC, Parks M. 2000. What is the true solubil-ity advantage for amorphous pharmaceuticals? Pharm Res17(4):397–404.
11. Craig DQ, Royall PG, Kett VL, Hopton ML. 1999. The rel-evance of the amorphous state to pharmaceutical dosageforms: Glassy drugs and freeze dried systems. Int J Pharm179(2):179–207.
12. Savolainen M, Kogermann K, Heinz A, Aaltonen J, PeltonenL, Strachan C, Yliruusi J. 2009. Better understanding of dis-solution behaviour of amorphous drugs by in situ solid-stateanalysis using Raman spectroscopy. Eur J Pharm Biopharm71(1):71–79.
13. Alonzo DE, Zhang GG, Zhou D, Gao Y, Taylor LS. 2010. Under-standing the behavior of amorphous pharmaceutical systemsduring dissolution. Pharm Res 27(4):608–618.
14. Slavina PA, Sheena DB, Shepherd EEA, Sherwood JN, FeederN, Docherty R. 2002. Morphological evaluation of the (-polymorph of indomethacin. J Cryst Growth 237–239:300–305.
15. Kiang YH, Yang CY, Staples RJ, Jona J. 2009. Crystal struc-ture, crystal morphology, and surface properties of an investi-gational drug. Int J Pharm 368(1–2):76–82.
16. Gu CH, Rao D, Gandhi RB, Hilden J, Raghavan K. 2005. Usinga novel multicompartment dissolution system to predict theeffect of gastric pH on the oral absorption of weak bases withpoor intrinsic solubility. J Pharm Sci 94(1):199–208.
17. Kostewicz ES, Wunderlich M, Brauns U, Becker R, Bock T,Dressman JB. 2004. Predicting the precipitation of poorly sol-uble weak bases upon entry in the small intestine. J PharmPharmacol 56(1):43–51.
18. Siepe S, Lueckel B, Kramer A, Ries A, Gurny R. 2006. Strate-gies for the design of hydrophilic matrix tablets with controlledmicroenvironmental pH. Int J Pharm 316(1–2):14–20.
19. Sugawara M, Kadomura S, He X, Takekuma Y, Kohri N,Miyazaki K. 2005. The use of an in vitro dissolution andabsorption system to evaluate oral absorption of two weakbases in pH-independent controlled-release formulations. EurJ Pharm Sci 26(1):1–8.
21. Silva Oliveira M, Miguel Agustinho SC, De Guzzi PlepisAM, Tabak M. 2006. On the thermal decomposition of dipyri-damole: Thermogravimetric, differential scanning calorimet-ric and spectroscopic studies. Spectrosc Lett 39(2):145–161.
22. Beten DB, Gelbcke M, Diallo B, Moes AJ. 1992. Interac-tion between dipyridamole and Eudragit S. Int J Pharm88(1–3):31–37.
23. Kostewicz ES, Brauns U, Becker R, Dressman JB. 2002. Fore-casting the oral absorption behavior of poorly soluble weakbases using solubility and dissolution studies in biorelevantmedia. Pharm Res 19(3):345–349.
24. Sugano K. 2009. A simulation of oral absorption using classi-cal nucleation theory. Int J Pharm 378(1–2):142–145.
25. Pakula R, Pichnej L, Spychala S, Butkiewicz K. 1977. Poly-morphism of indomethacin. Part I. Preparation of polymor-phic forms of indomethacin. Pol J Pharmacol Pharm 29(2):151–156.
26. Arnold YE, Imanidis G, Kuentz MT. 2011. Advancing in-vitrodrug precipitation testing: New process monitoring tools anda kinetic nucleation and growth model. J Pharm Pharmacol63(3):333–341.
27. Adhiyaman R, Basu SK. 2006. Crystal modification of dipyri-damole using different solvents and crystallization conditions.Int J Pharm 321(1–2):27–34.
28. Avdeef A, Bendels S, Tsinman O, Tsinman K, Kansy M. 2007.Solubility–excipient classification gradient maps. Pharm Res24(3):530–545.
29. Berbenni V, Marini A, Bruni G, Maggioni A, Cogliati P. 2002.Thermoanalytical and spectroscopic characterization of solidstate dipyridamole. J Ther Anal Calorim 68(2):413–422.
30. Crowley KJ, Zografi G. 2003. The effect of low concentra-tions of molecularly dispersed poly(vinylpyrrolidone) on in-domethacin crystallization from the amorphous state. PharmRes 20(9):1417–1422.
31. Taylor LS, Zografi G. 1997. Spectroscopic characterization ofinteractions between PVP and indomethacin in amorphousmolecular dispersions. Pharm Res 14(12):1691–1698.
32. Patterson JE, James MB, Forster AH, Lancaster RW, ButlerJM, Rades T. 2007. Preparation of glass solutions of threepoorly water soluble drugs by spray drying, melt extrusionand ball milling. Int J Pharm 336(1):22–34.
33. Patterson JE, James MB, Forster AH, Lancaster RW, ButlerJM, Rades T. 2005. The influence of thermal and mechanicalpreparative techniques on the amorphous state of four poorlysoluble compounds. J Pharm Sci 94(9):1998–2012.
35. Madieh S, Simone M, Wilson W, Mehra D, Augsburger L.2007. Investigation of drug–porous adsorbent interactions indrug mixtures with selected porous adsorbents. J Pharm Sci96(4):851–863.
36. Shibata Y, Fujii M, Kokudai M, Noda S, Okada H, Kondoh M,Watanabe Y. 2007. Effect of characteristics of compounds onmaintenance of an amorphous state in solid dispersion withcrospovidone. J Pharm Sci 96(6):1537–1547.
JOURNAL OF PHARMACEUTICAL SCIENCES DOI 10.1002/jps
