Results and discussion
The weights of the samples showed no randomness because the contributors to the total sample weight (i.e., drug weight, capsule
weight, and filler weight) were fixed with fairly little variation from the mean of the individual parameters. This uniformity
was reflected by the low percent RSD (0.49–1.31%) that indicated that the process of encapsulation had been accurate and efficient.
Rofecoxib had a higher overall weight because the original average weight (200 mg) of the tablets was higher than that of
propranolol (110 mg).
Disintegration. A review of the video of the process of capsule disintegration did not reveal any differences between DBcaps and standard
gelatin capsules. On average, the capsules ruptured after 60 s; the range of the lag was 34–70 s. Some capsules did not rupture;
instead, the capsule cap separated from the body, thus exposing the content to the medium. The individual parameters (i.e.,
capsule type, drug type, and filler) showed no effect on D-time (see Table I). The effects of interactions bewteen parameters
on disintegration time also were not significant.
The D-time for rofecoxib was slightly higher than that for propranolol, possibly because of its lower solubility; the average
D-times were 3.82 ± 0.28 min and 3.69 ± 0.44 min, respectively. Comparing the capsule effect, the average D-times were 3.85
± 0.28 min and 3.67 ± 0.43 min for DBcaps and standard gelatin capsules, respectively. These differences were not statistically
significant, however (p = 0.05). In the case of the filler, average D-times were 3.75 ± 0.36 min and 3.77 ± 0.39 min for the presence and absence
of filler, respectively. These times indicated that the filler had no effect on disintegration time.
As expected, the D-time for encapsulated tablets was greater than that for plain tablets. D-times for unencapsulated tablets
were 1.80 and 1.93 min, compared with 3.69 min and 3.82 min for the overencapsulated propranolol and rofecoxib tablets, respectively.
This difference resulted from the time lag required for the capsules to rupture and expose their content to the disintegration
medium. All the samples passed USP specification of D-time of less than 30 min. The residual plot (see Figure 1) was random without any trend, thus indicating
that the parameters had no effect on disintegration time.
Figure 1 (FIGURES ARE COURTESY OF THE AUTHORS)
Dissolution. All the six Delphian fiber-optic probes had good linearity (the range of R
was 0.986–0.998, where R
was the coefficient of determination) for propranolol and rofecoxib. The time taken for 80% of the drug to dissolve (T
80) also was used to compare the results of dissolution profiles statistically. The significance-of-effects table (Table II)
showed that the drug had an effect on T
80 (p < 0.0001). Although the interaction between drug and capsule appeared significant (p = 0.0219), this significance is of no consequence for formulation because the drugs belong to two different BCS classes.
The effects of the filler (p = 0.5716) and type of capsule (p = 0.7614) were not significant. The average T
80 times were 14.39 min and 13.80 min for presence and absence of filler, respectively. The differences were not statistically
significant. The same weight of backfill was used for all capsule types although the DBcaps size B capsules were smaller in
volume than HGC size 00 capsules. Differences in T
80 for the two capsule types were not significant, thus the volume of the fill did not have an effect on dissolution profile.
Similarity factor (
). The dissolution profiles were also compared using the f
2 similarity factor. A lag time of 2 min was observed in the dissolution profiles of encapsulated and unencapsulated tablets,
and it was incorporated into f
2 computations (13). An f
2 greater than 50 indicates that two dissolution profiles are similar. Propranolol attained a dissolution plateau faster than
rofecoxib because of differences in the drugs' solubilities in the dissolution medium. For propranolol tablets, the f
2 between standard gelatin and DBcaps capsules was 60.67, thus indicating that the dissolution curves of the two were similar
(see Figure 2). The f
2 between unencapsulated propranolol tablets and DBcaps capsules was 59.48. The f
2 between unencapsulated propranolol tablets and standard gelatin capsules was 53.94. The similarity factor between rofecoxib
tablets encapsulated in DBcaps and HGC capsules was 62.48 (see Figure 3). The f
2 between unencapsulated rofecoxib tablets and DBcaps capsules and that between unencapsulated rofecoxib tablets and standard
gelatin capsules were 56.97 and 54.92, respectively.
After storing the encapsulated tablets at room temperature and humidity conditions for one year, the authors found no differences
in the dissolution profiles of propranolol and rofecoxib tablets encapsulated using DBcaps or standard gelatin capsules. This
observation implies that the double-walled DBcaps capsules could be used for blinding drug candidates for a clinical study.
Although encapsulation resulted in a lag time of 2–3 min in disintegration compared with the unencapsulated tablets, the disintegration
and dissolution of propranolol and rofecoxib were the same whether tablets were encapsulated in DBcaps capsules or standard
gelatin capsules. It can thus be expected that in vitro drug release will not be influenced by the type of capsule used for overencapsulation in the double-blind clinical study.
The bioequivalence between the unencapsulated and encapsulated tablet is not inferred in this study; however, the in vivo disintegration data of unencapsulated and overencapsulated tablets reported by Wilding et al. suggest that the in vitro disintegration lag time is negligible in vivo (2).
Nevertheless, it is important to use DBcaps capsules for test and comparator dosage forms to correlate in vitro data, especially for the highly soluble drugs (e.g., those of BCS Class I). The presence or absence of backfill had no significant
effect on disintegration or dissolution. In effect, capsules offer an easy and inexpensive way of blinding clinical trials
and pose no threat to studies that compare the performance of two drugs. DBcaps-type capsules allow for the overencapsulation
of tablets with large diameters and make blind breaking in clinical trials virtually impossible.
The authors acknowledge Capsugel for funding the project, Delphian for the Delphian fiber-optic dissolution-monitoring system,
and Frank D'Amico, professor of statistics at Duquesne University, for assisting with data analysis.
Fredrick Esseku is a graduate student at the School of Pharmaceutical Sciences, Duquesne University. Mark Lesher is a pharmacist at Hershey Medical Center. Vishal Bijlani is a formulation- and process-development scientist at Patheon Pharmaceuticals. Samantha Lai is chief executive officer of Tamlylin Consultant and Contract Services. Ewart Cole is a consultant to the Capsugel division of Pfizer. Moji Christianah Adeyeye* is a professor of pharmaceutics and manufacturing science at Duquesne University, 441 Mellon Hall, School of Pharmacy, Duquesne
University, Pittsburgh, PA 15282, tel. 412.396.5133, email@example.com
*To whom all correspondence should be addressed.
Submitted: July 24, 2009. Accepted: Sept. 16, 2009.
1. S. Talpes et al., Int. J. Clin. Pharmacol. Ther.
43 (1), 51–56 (2005).
2. I.R.Wilding et al., J. Clin. Pharmacol.
45 (1), 101–105 (2005).
3. E. Fuseau et al., Clin. Ther.
23 (2), 242–251 (2001).
4. Y. Wu, F. Zhao, and M. Paborji, Pharm. Dev. Technol.
8 (4), 379–383 (2003).
5. J. Hogan et al., Pharm. Res.
13 (6), 944–949 (1996).
6. M. Bartolomei et al., J. Pharm. Biomed. Anal.
21 (2), 299–309 (1999).
7. P. Modamio et al., Int. J. Pharm.
130 (1), 137–140 (1996).
8. B. Mao et al., J. Pharm. Biomed. Anal.
28 (6), 1101–1113 (2002).
9. J.W. Moore and H.H. Flanner, Pharm. Technol.
20 (6), 64–74 (1996).
10. FDA, Dissolution Testing of Immediate-Release Solid Oral Dosage (Rockville, MD, Aug. 1997).
11. V.P. Shah et al., Pharm. Res.
15 (6), 889–896 (1998).
12. V. Bijlani et al., AAPS PharmSciTech.
8 (3), pp.1–4 (2007).
13. National Institute of Health Sciences, Guidance for Bioequivalence Studies for Different Strengths of Oral Solid Dosage Forms (Tokyo, 2000).