The influence of monoglycerides, mixtures and derivatives on the formation of pellets by extrusion/spheronization

In a recent publication,¹ we have shown that it is possible to incorporate the glyceride, glycerly monostearate (GMS), into pellet formulations at a level of at least 30% and still produce round pellets, without reducing the in vitro drug release. The formulations required less water than those containing microcrystalline cellulose (MCC) alone as the spheronization aid, which appears to be the reason for the improved stability of formulations containing ranitadine hydrochloride.²

The inclusion of GMS also offers the opportunity to incorporate drugs that are water insoluble or are poorly permeable through the gastrointestinal membrane into the nonaqueous phase of the formulation.

Table 1. Formulation codes for the additives to the standard formulation of 10% diclofenac sodium, 30% additive, with the remainder of the formulation being MCC.

GMS is not a pure material — even the Japanese Pharmacopoeia (JP), European Pharmacopoeia (EP) and the United States National Formulary (USNF) differ in their standards.³ The question arises, therefore, do these allowable differences change the performance of the products, and would other types of glycerides used in pharmaceutical formulations (e.g., as absorption enhancers) perform in the same way? To investigate these issues, we have used a previously reported successful formulation that contained a model drug (10% diclofenac sodium) with a range of glycerides as an additive (30%), MCC (60% ) and water. As a comparison, a formulation containing 10% drug with 90% MCC is reported as a standard. The glycerides are glyceryl monostearate of differing composition; a mixture of mono and diglycerides; and hydrogenated natural oil glycerides.

Materials and methods Materials

The range of compounds studied was

• GMS (lot number 209215; Huls [UK] Ltd), containing 40–50% GMS, melting range of 57–59 °C, EP grade.

• Imwitor 900 (Huls) containing 40–50% monoglycerides with a melting range of 56–61 °C.

• Imwitor 191 (Huls) containing a minimum of 90% monoglycerides, mainly GMS and glyceryl monopalmitate, melting range of 66–71 °C.

• Imwitor 742 is described as a medium chain (caprylic/capric) partial glyceride with a minimum monoglyceride content of 45%, melting range of 22–27 °C.

• Glycerol esters of hydrogenated coglycerides (Softisan 142), with a melting range of 42–44 °C, and hydrogenated palm oil (Softisan 154), with a melting range of 53–58 °C, were also supplied by Huls.

The GMS and Imwitor 900 were ground in a small coffee grinder (Bruan AG, Germany) and passed through a 250 μm aperture sieve. The Softisans were again ground in the coffee grinder, but a 500 μm aperture sieve was used because of the greater difficulty in grinding the softer materials. Imwitor 191 and 742 were used as received. Imwitor 724 was soft enough to be incorporated directly into the powder when preparing the wet powder mass.

MCC was Avicel PH101 (batch 6521; FMC International, Ireland). The model drug (diclofenac sodium) was purchased from Profarma Nobel (Sweden) and had a mean Feret's number diameter of 58.4 μm.

The materials were combined in the ratio 10% drug, 30% additive and 60% MCC, and given the formulation code set out in Table 1. Where no additive is recorded, the MCC content was increased to 90%.

Table 2 . Particle size, apparent particle density and contact angle of the materials used in the pellet formulations (values ± standard deviation).

Methods

The particle size of the materials tested in the formulations was determined by an image analyser (Seescan Solitaire 512; Seescan, UK), connected to a black and white camera (CCD-4 miniature video camera; Rengo Co. Ltd, Japan) and an Olympus BH-2 microscope (Olympus, Japan), set at a suitable magnification to the size of the particles being measured. The Feret diameter, a mean distance between two parallel tangents (obtained from 36 measurements around a particle in 10 ° increments) of 500 particles, was obtained and the number average particle size was determined from Edmonson's equation 4 (Table 2). The apparent particle density of the raw materials and the apparent pellet density were determined with an air comparison pycnometer (Model 930; Beckman, USA) using ambient air. Three volume measurements of known sample weights were made and the values for the raw materials given in Table 2.

Table 3 . The quantity of water required, extrusion force and type of flow for various formulations extruded under standard conditions.

The value for the contact angles of water on the glycerides was determined with a Cahn dynamic contact angle analyser (Cahn Instruments, USA) using the Wilhemly plate method.⁵ Plates of the materials were prepared by compacting rectangular samples of the powders in a specially prepared punch and die set (20 mm×20 mm) using a Specac hydraulic press (Specac Ltd, Kent, UK).

For Imwitor 724, a glass plate was covered with the material by immersing it in a molten sample of the material and cooling slowly at room temperature. The perimeter of each plate was measured with a micrometer+10 μm. The water used was freshly prepared distilled and the values in Table 2 are the mean of five sample plates.

The pellets were prepared by first mixing the powders in a planetary mixer, adding the water, extruding the wet mass at 200 mm/min through a 1 mm diameter die attached to a 2.54 cm diameter barrel of a ram extruder, followed by spheronization on a 12 cm diameter crosshatched plate (Caleva, UK), rotating at 1800 rpm for 10?min. The full details of the whole process have been described previously.⁶ The pellets, except those containing Imwitor 742, were dried at 40 °C for 14 h in a hot air oven (Pickstone Ltd, UK). Pellets containing Imwitor 742 were allowed to dry at room temperature until a constant weight was achieved. This usually took 48 h.

The particle size of the pellets was determined by sieve analysis using a set of British Standard sieves in a square root 2 progression of sieve sizes, with the introduction of a 1700 μm sieve between the 1400 μm and 2000 μm sieves to provide a more detailed analysis. From the cumulative per cent undersize graph, the median and the interquartile range of the size distribution were obtained, while the mode and the percentage of pellets in the modal size were obtained from the sieve analysis. The shape factor (eR) was obtained⁷ by the same image analyser used to determine the particle size of the powders. The porosity of the pellets was obtained as 1 minus the ratio of the apparent pellet density to the apparent particle density of the powdered components of the pellets.

The drug release from the pellets was assessed by USPXXXI dissolution apparatus II or paddle method (Pharmatest Dissolution Tester, Type PTWS, Germany), with a paddle speed of 100 rpm in 900?mL of simulated intestinal fluid (without pancreatin) at pH 7.5, with the quantity of pellets chosen to ensure that sink conditions were attained. The quantity of diclofenac in the samples, which were taken automatically at known time intervals, was analysed by UV spectrophotometry (UV-Vis spectrophotometer Model 554; Bodenseewerk Perkin-Elmer & Co. GmbH, Germany). Six beakers were used for each pellet formulation.

Results and discussion

The water level in 60% of the dry solid content, which had been shown to be suitable for the GMS preparation G01,¹ was selected as the starting point for the other glycerides. It was found to be appropriate for the range of other glycerides, except for the Imwitor 742. Here, the water required to provide a smooth extrudate and round pellets with a narrow size distribution had to be reduced to just 16% (Table 3). This glyceride has the lowest melting range and would be the most deformable of the glycerides, although with the level of water added, it required the greatest force to extrude the mass. In all cases, the extrudate was completely smooth and the extrusion displacement gave a constant steady state extrusion force.

Table 4 Particle size characteristics of the pellets produced from the various formulations.

All the extrudates from the different formulations rounded to form pellets within the 10 min of the spheronization process. The modal size fraction, the percentage in this fraction, the median size and the interquartile range show that there is a good uniformity in the pellets produced from the different formulations (Table 4). In particular, the three types of GMS produce pellets that are very similar and thus the differences in the types is not important to their function. As observed previously, the size is larger than pellets prepared with just MCC. The two hydrogenated glycerides, Softisan 142 and 154, produce pellets that are very similar in size and size distribution, but the inclusion of Imwitior 742 produces larger pellets with a similar size range.

Figure 1 The MDT as a function of the contact angle of the glyceride (linear correlation coefficient r = -0.843)

When the shape of the pellets is considered, it is clear that all the various formulations produce pellets that are acceptably spherical — eR values greater than 0.58 and aspect ratios less than 1.01 (Table 5). The least round pellets were those containing Imwitor 742, which had the highest extrusion force and the largest pellet size. The porosity of the pellets varied with the formulation. The highest porosity was for the grade of GMS containing 40–50% GMS, which would have been expected to be very similar to batch G01. The source of this difference is difficult to explain as pellets containing these two grades were very similar in all other respects. The pellets with the lowest value for porosity were prepared from the formulation containing Imwitor 742, the material that showed the greatest differences in other properties.

Table 5 Shape factor eR, aspect ratio, apparent pellet density and porosity of pellets prepared form the various formulations (mean values ±standard deviations; n=50 for eR and aspect ratio; n=3 for apparent pellets density and porosity).

The in vitro drug release, characterized by the use of statistical moments,⁸ is presented in Table 6. All formulations containing the various glcyerides release the drug faster than the formulation containing only MCC and drug. There appears to be no consistent release mechanism controlling the process, which is perhaps not surprising as the time period is relatively short — representing an immediate release rather than an extend release profile. Also, in the case of all the pellets containing the various glycerides, fracturing of the surface during the dissolution test was observed, indicating that there was not a constant surface area from which the dissolution occurred

It is difficult to represent the physical properties of the glycerides to be able to relate their properties to those of the pellets. As they are all relatively hydrophobic, the contact angle may be a suitable measure. When the formulation/pellet properties of quantity of water required, mean steady state extrusion force, median pellet diameter and pellet porosity were plotted as a function of the contact angle, none of the graphs indicated a potential relationship. When, however, the mean dissolution time (MDT) was plotted as a function of the contact angle, there was a clear indication that the value of the contact angle increased as the value of MDT decreased (Figure 1). This appears contradictory, but implies that the presence of the hydrophobic material does not inhibit ingress of fluid into the pellet and in fact appears to assist the ingress of active fluid. The aqueous fluid that penetrates the MCC appears to be held firmly and delays the release of the drug from the pellet structure.

Table 6 Dissolution characteristics of the various pellet formulations. MDT; variance associated with the MDT; relative dispersion of the dissolution time (RD) and the release model of the drug release from the pellets, 1, first order; 2, square root; N, no simple kinetics. The pellet size was the 1400–1700 μm size fraction and the values are the mean 6 the standard deviations of ± replicates.

Conclusions

The results show that variability in the type of GMS does not alter the ability of this additive to perform its function as an enhancer of pellet formation by extrusion/spheronization. For three different types, the water level required to provide smooth regular extrudate at equivalent extrusion forces, which readily form spherical pellets, was the same and is considerably less than that required with MCC alone. The only difference in performance was that one batch of the GMS (Imwitor 191) produced pellets that were more porous, but the size, shape and drug release performance was the same. It was possible to replace the GMS with hydrogenated glycerides and retain the pellet performance. All these formulations offer the opportunity to reduce the water required to form the pellets, which may have implications for the stability of some drugs. The replacement of the GMS with Imwitor 742 allowed an even further reduction in the quantity of water required for pellet formation. The fastest drug release was obtained with the additive Softisan?142, which was the most hydrophobic, having the highest value for the contact angle against water.

Thus, pharmaceutically acceptable glycerides provide an additional option when it comes to the formulation of pellets by extrusion/spheronization.

References

1. J. Chatchawalsaisin, F. Podczeck and J.M. Newton, Eur. J. Pharm. Sci.24(1), 35–48 (2005).

2. A.W. Basit, J.M. Newton and L.F. Lacey, Pharm. Dev. Tech.4(4), 499–505 (1999).

3. R.C. Rowe, P.J. Sheskey and P.J. Weller, Handbook of Excipients 4th Edition

(Pharmaceutical Press, London, UK, 2003).

4. I.C. Edmonson, "Particle Size Analysis" in H.S. Bean et al. Eds., Advances in Pharmaceutical Sciences Volume 2 (Academic Press, London, UK, 1967) pp 95–179.

5. A.E. Alexander and P. Johnson, Colloid Science (Oxford University Press, London, UK, 1949) pp 490–492.

6. J.M. Newton et al., Pharm. Tech. Eur.16(10), 21–27 (2004).

7. F. Podczeck and J.M. Newton, J. Pharm. Pharmac.46(2), 81–85 (1994).

8. D.Veogle, D.Brockmeier and H.M. von Hattingberg, "Modelling of Input Function to Drug Absorption by Moments," Proceedings of the Symposium on Compartmental

and Non-Compartmental Modelling in Pharmacokinetics" (Smolenice, The Czeck Republic, 12–16 September 1994) pp1–14.

9. S. Boutel, "Factors Influencing the Preparation of Spherical Granules by Extrusion/Spheronization," PhD Thesis University of London, London, UK (1995).

J. Michael Newton is Emeritus professor of pharmaceutics at the School of Pharmacy, University of London, UK.

Jittima Chatchawalsaisin is a lecturer in pharmacy at the Faculty of Pharmaceutical Sciences, Chulalangkom University, Thailand.

Fridrun Podczeck is professor of pharmaceutics at Sunderland School of Pharmacy, University of Sunderland, UK.