Professional Documents
Culture Documents
Drug Product Science and Technology, Bristol-Myers Squibb, 1 Squibb Drive, New Brunswick, New Jersey 08901
ABSTRACT: The objective of this study was to establish the effects of the level of minor formulation components (sodium lauryl sulfate:
SLS, and magnesium stearate: MgSt) and manufacturing process on final blend compaction properties and the performance of the tablets
during film coating. A 2 × 2 × 3 factorial study was conducted at two levels of SLS (0% and 1%, w/w) and MgSt (0.5% and 1.75%,
w/w), along with three different manufacturing processes (direct compression, high-shear wet granulation, and dry granulation). The tablets
were compressed to the same solid fraction (0.9) and the resulting tablet hardness values were found to vary over a range of 13–42
SCU, highlighting large compactability differences among these batches. Increase in the level of SLS or MgSt in the formulation had a
significant negative effect on compactability and the performance of film-coated tablets. The detrimental effects on compaction and coating
performance were magnified for the dry granulation process, likely due to the overall increased shear experienced by excipients (SLS,
MgSt, microcrystalline cellulose) during the roller compaction and milling steps. The findings of this study highlight the importance of the
manufacturing process when considering the use-level of formulation components such as SLS and MgSt in the formulation. C 2014 Wiley
Periodicals, Inc. and the American Pharmacists Association J Pharm Sci 103:3666–3674, 2014
Keywords: excipients; coating; compression; processing; powder technology
Table 1. Experimental Design Details on the Split of SLS and MgSt Between the Intra- and Extragranular Portions for the Three
Manufacturing Processes
Batch # 1 2 3 4 5 6 7 8 9 10 11 12
Intragranular portion
MgSt – – – – 0 0 0 0 0.25 0.25 0.75 0.75
SLS – – – – 0 1 0 1 0 1 0 1
Extragranular portion
SLS 0 1 0 1 0 0 0 0 0 0 0 0
MgSt 0.5 0.5 1.75 1.75 0.5 0.5 1.75 1.75 0.25 0.25 1 1
Total
SLS 0 1 0 1 0 1 0 1 0 1 0 1
MgSt 0.5 0.5 1.75 1.75 0.5 0.5 1.75 1.75 0.5 0.5 1.75 1.75
to be quite hydrophobic in nature, which in turn results in a of these excipients are shown in Table 1 and explained in the
loss of adhesion between the tablet substrate and the coating subsequent manufacturing process section. Statistical regres-
film.14–18 sion analysis was conducted using JMP statistical software
R
There are two important aspects in this field that have not (SAS).
been explored, and form the motivation for the current work: (1)
interaction of the effects of SLS and MgSt with the manufactur- Manufacturing Process
ing process, and (2) the impact of manufacturing process and The three manufacturing processes used in this study are de-
SLS level on tablet coating performance. With this knowledge- scribed below (Fig. 1). The blending conditions such as, blender
gap in mind, the objective of this work is to establish the main scale, blending time, and blender speed were maintained con-
effects and interactions of manufacturing process, SLS level, stant across all batches to eliminate any blending-related dif-
and MgSt level on the final blend compaction properties and ferences in blend lubrication. Final blends from each of the
performance of the core tablets during the film coating. A DOE manufacturing processes were compressed into oval-shaped
(design of experiments) study was conducted using two levels tablets at a press weight of 200-mg using embossed punches.
of SLS (0% and 1%, w/w), two levels of MgSt (0.5% and 1.75%, These oval-shaped tablets were used for the coating perfor-
w/w), and three different manufacturing processes (DC, WG, mance evaluation. For film coating studies, the tablets were
DG). The responses of the DOE study included granule com- compressed to a solid fraction of 0.9. Compaction profiles (over
paction properties and performance of core tablets during film a range of solid fractions) using flat-faced punches were gen-
coating as measured by logo bridging coating defects. erated using a Stylcam compaction simulator (Medel’Pharm,
R
1.75%, w/w), and three different manufacturing processes (DC, 197S (Quadro Engineering, Waterloo, Ontario) running at 1300
WG, DG). The split of SLS and MgSt between the intragran- rpm using a 8 mm opening screen and dried in a 25-L Freund
ular and extragranular portions (for the DG and WG batches) Vector fluid bed at an inlet temperature of 50◦ C to a target LOD
Figure 1. Schematic representation of the three manufacturing processes: (a) direct compression, (b) wet granulation, and (c) dry granulation.
(loss on drying @ 105◦ C) value of less than 3.0% (w/w). The dry into the granulator. The granulator mill was fitted with a 3.15
granules were then passed through a Quadro Comil running
R
mm upper screen and 1.00 mm lower screen, and operated at
at 2300 rpm using a 1.1 mm screen opening. Extragranular 80 rpm. The granulated material was placed into a 25-L bin
croscarmellose sodium was mixed with the milled granules in a blender and extragranular excipients were added. Croscarmel-
bin blender at 25 rpm for 75 revolutions followed by lubrication lose sodium was added and blended at 25 rpm for 75 revolutions
with MgSt (prescreened using 30-mesh) for 100 revolutions at followed by lubrication with MgSt (prescreened using 30-mesh)
25 rpm. A schematic diagram of the WG process is shown in for 100 revolutions at 25 rpm. A schematic diagram of the DG
Figure 1b. process is shown in Figure 1c.
paction process at a 4-kg scale. SLS was added along with with 1 spray gun, to a target theoretical weight gain of 4.0%
anhydrous lactose, microcrystalline cellulose, and croscarmel- (w/w) at a batch size of 800 g. A coating suspension at 15% (w/w)
lose sodium to a 25-L bin blender and blended at 25 rpm solids was used. Coating process conditions were kept constant
for 263 revolutions. The intragranular portion of MgSt was across all batches and were monitored using PyroButton data R
screened using a 30-mesh screen, prior to blending for 125 revo- loggers (Opulus, Philadelphia, PA). Information on PyroBut-
lutions at 25 rpm. The resulting preblend was granulated using tons and their relevance to the coating process measurements
the Alexanderwerk WP-120 (Montgomeryville, Pennsylvania)
R
in terms of logo bridging tablet defects has been established
roller compactor equipped with a smooth upper roll and knurled by the authors in other publications.18–20 Trial experiments
bottom roll. A roll gap size of 2.2 mm was targeted to achieve a were conducted to estimate the coating conditions required to
ribbon density in the range of 1.00–1.20 g/cc. This was achieved achieve a wide range of logo bridging responses across the 12
by adjusting the screw speed (8–12 rpm) and roll pressure batches. PyroButton data showed that a tablet-bed RH of 40%
(45 ± 5 bar). The ribbons were crushed using the mill integrated showed no logo bridging for the DG batch with low SLS and
MgSt level combination and 100% logo bridging for the high
SLS and MgSt level combination batch. Therefore, those pro-
cess parameters were selected for this study. The selected pro-
cess parameters selected were: a target spray rate of 7.5 g/min,
inlet airflow of 40 cfm, exhaust temperature of 43◦ C, pan speed
of 25 rpm, and a gun-to-bed distance of 3.75 in. The coater was
operated in “exhaust temperature-control” mode, meaning that
the unit self-adjusted the inlet air temperature to reach and
maintain the exhaust temperature at the set value.
Physical Characterization
True Density
True density of final blend was measured by AccuPyc II 1340
R
Compaction
least one of the four letters of the logo on a tablet face was
Compaction data were generated using Stylcam compaction
R
Table 2. Tablet Hardness Values for the 12 DOE Batches at the Same Tablet Solid Fraction of 0.9, as a Measure of the Compaction
Differences Between These Formulations
Formulation
Change in Hardness w.r.t.
Manufacturing Process SLS (%, w/w) MgSt (%, w/w) Hardness, SCU Reference Batch (%)
Figure 3. Compaction profiles for the wet granulated batches. Error Figure 4. Compaction profiles for the roller compacted batches. Error
bars indicate one standard deviation values (n = 3 tablets). bars indicate one standard deviation values (n = 3 tablets).
MgSt is known to be an effective lubricant which means that were in the order of batch # 9>11>10>12. For DG, a more sig-
it coats the other particle during the blending process, result- nificant drop in tablet hardness (an indicator of compactabil-
ing in a reduced bonding between these particles.21,22 This can ity) was observed when using either 1% SLS (37% reduction,
explain the significant loss in compaction when MgSt is added Table 2), or 1.75% MgSt (43% reduction), or both in combina-
to a formulation. SLS also has some lubricant-like properties tion (56% reduction), as compared with the reduction in hard-
and works in a similar way resulting in a loss of compaction ness values observed for DC and WG for similar composition
when blends are compressed into tablets.5,23 It is worthwhile changes. It should be noted that a portion of MgSt and all of SLS
noting that the loss in compaction as a result of increase in was added intragranularly in case of the DG process. The higher
SLS amount alone (batch # 2 vs. # 1) was similar in magnitude observed reduction in compaction for the DG batches compared
to the loss of compaction due to the increase in MgSt amount with the other manufacturing processes is attributed to addi-
(batch # 3 vs. # 1). tional shear imparted on the blend by the roller compactor and
milling of the ribbons. These additional processesing steps can
Wet Granulation result in an enhanced negative effect of SLS and MgSt on com-
The tabletability data of the four wet granulated batches paction. A combined use of SLS and MgSt in this case led to
(# 5–8) are shown in Figure 3. The tabletability of the batches extremely poor compaction, and as a result, the lowest tablet
were in the order of batch # 5>7>6>8. Table 2 shows a com- hardness among all batches in the study.
parison of the tablet hardness values across these batches at a When comparing the three manufacturing processes with
solid fraction of 0.9. each other it must be kept in mind that the other formula-
An increase in SLS level from 0% to 1% (w/w) was observed tion excipients such as microcrystalline cellulose undergo dif-
to cause a greater loss in compaction than an increase in MgSt ferent amounts of processing and this will also contribute to
level from 0.50% to 1.75%. This trend was opposite to what the observed overall compaction properties of the final blend.25
was observed for the DC batches (Fig. 2). The reversal in trend However, when comparisons are made to the reference batches
is likely due to the differences in mode of addition of SLS be- within the same manufacturing process, those effects are nor-
tween the DC and WG batches. For the WG batches, the SLS malized.
was added in the intragranular portion causing it to undergo
additional high-shear mixing and compiling steps. It has been Statistical Analysis
shown in the literature that additional shear imparted to SLS Statistical analysis of the compaction data show that all the
can be detrimental to tablet hardness.5 SLS itself being a lu- three main factors (manufacturing process, SLS, and MgSt
bricant is observed to be exhibiting similar properties to that concentration) studied are significant (p < 0.05). The output
of MgSt. MgSt has a plate-like structure that tends to shear variable used in the statistical model was the hardness value
apart during processing (blending or any activities that tend to of the tablet at 0.9 solid fraction. RSD on tablet hardness val-
apply shear), thereby causing an increased effective concentra- ues was in the range of 3%–5%. Table 3 shows a summary of
tion of MgSt that tends to coat more particles and resulting in
further reduction in interparticle bonding.10,11,24 It is observed
that SLS shows a similar behavior with increased processing Table 3. Regression Analysis for the DOE with Tablet Hardness as
causing a more pronounced effect on the loss in compaction. the Response
Additionally, the addition of MgSt alone (batch # 7 vs. 5#) did Term Estimate Standard Error t Ratio Prob > t
not show the same extent of reduction in hardness (Table 2) as
was observed for the DC case (batch # 3 vs. 1). Intercept 37.53 1.89 19.88 0.0000
Process (DC) − 0.17 1.17 − 0.14 0.8904
Dry Granulation Process (RC) − 6.42 1.17 − 5.50 0.0009
MgSt − 6.40 1.32 − 4.85 0.0019
The tabletability data of the four dry granulated batches (# 9– SLS − 8.33 1.65 − 5.05 0.0015
12) are shown in Figure 4. The compactability of the batches
Figure 7. Logo bridging coating defect response as a function of MgSt and SLS concentration when grouped by the manufacturing process.
logo bridging. A low level of MgSt with no SLS (batch # 1, 5, and observed logo bridging. This is in agreement with reported lit-
9), did not show any logo bridging for any of the manufactur- erature on MgSt where it was shown that a higher amount of
ing processes. The worst logo bridging was observed in batches MgSt results in an increase incidence of logo bridging or poor
containing high level of MgSt that were roller-compacted. At adhesion between the film and tablet substrate.16 Lubricants
the higher MgSt amount, the effect of adding 1% SLS was ob- coat the granule particles and present a tablet surface that con-
scured since both batches (batches #11, 12) resulted in 100% sists mainly of non-polar hydrocarbon groups, thereby lowering
logo bridging. adhesion between the polymer (HPMC in this case) in the film
and the tablet substrate.16
In this study, the addition of 1.75% (w/w) MgSt alone did
Statistical Analysis
not affect the number of logo bridging tablet defects for DC or
Statistical analysis of the coating performance data reveal that WG cases (Fig. 7). This indicates that the degree of lubrication
all the three main factors (manufacturing process, SLS, and achieved for these cases was not high enough to cause a logo
MgSt concentration) studied are significant (p < 0.05). Table 4 bridging tablet defect, even though it may have affected the
shows a summary of the regression analysis. The output vari- adhesion between the film and the substrate to a certain extent.
able used in the statistical model was the% logo bridging of There was an indication for the DC and wet granulated batches
the coated tablets. The analysis showed a p value of 0.0001 for that 1% (w/w) SLS in the tablet formulation had a negative
the manufacturing process, 0.0001 for MgSt concentration, and effect on the performance during coating (Fig. 7). This effect
0.03 for SLS concentration, with process and MgSt interaction was prominent for the roller compacted batch with 1% (w/w)
term being significant at p value of 0.0001. SLS, with 22% of the batch showing logo bridging. For all three
A cross-sectional view of a tablet (sliced) showing logo bridg- manufacturing processes, a combined use of 1% (w/w) SLS with
ing is shown in Figure 8, demonstrating the lack of adhesion 1.75% MgSt caused significant overlubrication to show-up as a
of film to the tablet substrate as the underlying reason for the logo bridging tablet defect, with a more pronounced effect in
Table 4. Regression Analysis for the DOE with % Logo Bridging as the Response
Figure 8. A tablet showing logo bridging as seen from (a) top, (b) cross-section (sliced). The logo 468X is not easily readable due to the lifting
of the film (lack of adherence) from the substrate as obvious from the cross-sectional view.
the DG process, where the entire batch showed logo bridging. especially for a DG process. It is interesting to note how these
To the author’s knowledge, this negative effect of SLS on the minor formulation components can play such a significant role
performance of tablets during coating is not previously reported in the overall manufacturing scheme.
in the literature. This is likely the result of coating of excipient Additionally, as shown in this study, the formulation com-
particles by SLS, thereby reducing adhesion between the film ponents can also dramatically influence the performance of
and the substrate, and is clearly observed when SLS (added tablets during a film coating process. Therefore, an understand-
intragranular) undergoes sufficient shear imparted during the ing and quantification of which formulation components may
DG process. However, other possible causes playing a role in play a role and the interplay of the tablet formulation with the
adhesion of HPMC film such as an interaction of the coating manufacturing process is important. Once such a link is es-
film with the anionic charge of SLS and/or hydrophobic portion tablished, a formulator can make an informed decision when
of SLS molecule can not be overruled and will require additional selecting a certain manufacturing process and/or a coating
mechanistic characterization studies. system.
In lieu of these findings, appropriate risk analysis must be
conducted when excipients such as MgSt and SLS are present
CONCLUSIONS in the formulation at high concentrations (typically >1% w/w)
and shear is being applied during manufacturing. For exam-
It is clear from this study that MgSt and SLS play an important ple, when considering addition of a milling step for preblend
role in both compaction and coating performance. The negative preparation (to get rid of potential API lumps), a risk to down-
influence of these excipients was shown to be a function of their stream performance (compaction and coating) must be carefully
concentration and the manufacturing process used. It should be evaluated.
pointed out that the purpose of the study was to assess differ-
ences between the manufacturing processes the way they are
typically utilized and not necessarily to compare them in abso-
ACKNOWLEDGMENTS
lute terms. The magnitude of the observed effects will also be a
function of the mechanical properties of the other formulation The authors would like to thank Elena Zour, Jatin Patel, and
components (e.g., brittle versus ductile). The risks associated Brian Zacour from Bristol-Myers Squibb for their help with this
with the use of these excipients must be carefully evaluated, work.
REFERENCES 14. Rowe RC. 1977. The adhesion of film coatings to tablet surfaces—
The effect of some direct compression excipients and lubricants. J
1. Martinello T, Kaneko TM, Velasco MVR, Taqueda MES, Consiglieri Pharm Pharmacol 29(12):723–726.
VO. 2006. Optimization of poorly compactable drug tablets manufac- 15. Rowe RC. 2007. Film/tablet adhesion, film thickness, in-
tured by direct compression using the mixture experimental design. ternal stresses and bridging of the intagliations—A unified
Int J Pharm 322(1–2):87–95. model with practical implications. Eur J Pharm Sci 30(3–4):236–
2. Hausman DS. 2004. Comparison of low shear, high shear, and fluid 239.
bed granulation during low dose tablet process development. Drug Dev 16. Rowe RC. 1983. The coating of tablet surfaces by lubricants as
Ind Pharm 30(3):259–266. determined by a film/tablet adhesion measurement. Acta Pharm Suec
3. Gift AD, Luner PE, Luedeman L, Taylor LS. 2009. Manipulating 20(1):77–80.
hydrate formation during high shear wet granulation using polymeric 17. Zuurman K, Van der Voort Maarschalk K, Bolhuis GK. 1999. Effect
excipients. J Pharm Sci 98(12):4670–4683. of magnesium stearate on bonding and porosity expansion of tablets
4. Jørgensen A, Rantanen J, Karjalainen M, Khriachtchev L, Räsänen produced from materials with different consolidation properties. Int J
E, Yliruusi J. 2002. Hydrate formation during wet granulation stud- Pharm 179(1):107–115.
ied by spectroscopic methods and multivariate analysis. Pharm Res 18. Pandey P, Ji J, Subramanian G, Gour S, Bindra DS. 2014.
19(9):1285–1291. Understanding the thermodynamic micro-environment inside a pan
5. Moore F, Okelo G, Colon I, Kushner Jt. 2010. Improving the hardness coater using a data logging device. Drug Dev Ind Pharm 40(4):542–
of dry granulated tablets containing sodium lauryl sulfate. Int J Pharm 548.
400(1–2):37–41. 19. Pandey P, Bindra DS. 2013. Real-time monitoring of thermody-
6. Aljaberi A, Chatterji A, Dong Z, Shah NH, Malick W, Singhal D, namic microenvironment in a pan coater. J Pharm Sci 102(2):336–340.
Sandhu HK. 2013. Understanding and optimizing the dual excipi- 20. Pandey P, Bindra DS, Felton LA. 2013. Influence of process pa-
ent functionality of sodium lauryl sulfate in tablet formulation of rameters on tablet bed microenvironmental factors during pan coating.
poorly water soluble drug: Wetting and lubrication. Pharm Dev Technol AAPS PharmSciTech 15(2):296–305.
18(2):490–503. 21. Shah AC, Mlodozeniec AR. 1977. Mechanism of surface lubrication:
7. Wang J, Wen H, Desai D. 2010. Lubrication in tablet formulations. Influence of duration of lubricant-excipient mixing on processing char-
Eur J Pharm Biopharm 75(1):1–15. acteristics of powders and properties of compressed tablets. J Pharm
8. Baichwal AR, Augsburger LL. 1988. Variations in the fric- Sci 66(10):1377–1378.
tion coefficients of tablet lubricants and relationship to their 22. Lakio S, Vajna B, Farkas I, Salokangas H, Marosi G, Yliruusi
physicochemical properties. J Pharm Pharmacol 40(8):569– J. 2013. Challenges in detecting magnesium stearate distribution in
571. tablets. AAPS PharmSciTech 14(1):435–444.
9. Kikuta J, Kitamori N. 1985. Frictional properties of tablet lubricants. 23. Sheskey PJ, Robb RT, Moore RD, Boyce BM. 1995. Effects of lubri-
Drug Dev Ind Pharm 11(4):845–854. cant level, method of mixing, and duration of mixing on a controlled-
10. Johansson ME, Nicklasson M. 1986. Investigation of the film for- release matrix tablet containing hydroxypropyl methylcellulose. Drug
mation of magnesium stearate by applying a flow-through dissolution Dev Ind Pharm 21(19):2151–2165.
technique. J Pharm Pharmacol 38(1):51–54. 24. Johansson ME. 1985. Influence of the granulation technique
11. Leinonen UI, Jalonen HU, Vihervaara PA, Laine ESU. 1992. Phys- and starting material properties on the lubricating effect of
ical and lubrication properties of magnesium stearate. J Pharm Sci granular magnesium stearate. J Pharm Pharmacol 37(10):681–
81(12):1194–1198. 685.
12. Patel S, Kaushal A, Bansal A. 2007. Lubrication potential of mag- 25. Bultmann JM. 2002. Multiple compaction of microcrystalline cel-
nesium stearate studied on instrumented rotary tablet press. AAPS lulose in a roller compactor. Eur J Pharm Biopharm 54(1):59–
PharmSciTech 8(4):57–64. 64.
13. Perrault M, Bertrand F, Chaouki J. 2011. An experimental inves- 26. He X, Secreast PJ, Amidon GE. 2007. Mechanistic study of the
tigation of the effect of the amount of lubricant on tablet properties. effect of roller compaction and lubricant on tablet mechanical strength.
Drug Dev Ind Pharm 37(2):234–242. J Pharm Sci 96(5):1342–1355.