Optimizing Compression Coating for Pulsatile Release Tablets

Submitted: January 21 2019    
Accepted: July 9 2019

The concept of pulsatile release, in which active ingredient is fully and quickly released after a predetermined lag time, is gaining popularity as a drug-delivery approach. Pulsatile release offers a way to counteract the “first-pass” effect, and to enable nocturnal dosing. This article summarizes research into optimizing the formulation of pulsatile release tablets that contain diclofenac sodium (DS), an anti-inflammatory used to treat rheumatoid arthritis. Studies focused on optimizing the formulation of a time-controlled tablet containing the active ingredient in its inner core, surrounded by a coating of hydrophilic polymers. 

As the fields of chronopharmaceutics and chronopharmacology (1) reveal more about the importance of circadian rhythms with respect to human physiology, disease state, and drug action, time-controlled pulsatile-release tablets are being considered as a desirable drug-delivery mode. Research discussed in this article focused on compression coating, which can involve direct compression of both the tablet’s core and its coat, obviating the need for separate coating processes and the use of coating solutions. Compression-coated formulations can be used to protect hygroscopic, light-sensitive, oxygen- or acid-labile drugs, or to separate incompatible drugs from each other. They can also be used to achieve controlled release, and a number of studies (2) have evaluated compression-coated time-controlled drug delivery systems. Most of these formulations release the drug after a lag phase. 

Compression is easy to achieve on a laboratory scale. The technique requires relatively large amounts of coating materials, however, and it is difficult to position the cores correctly (3).  For pulsatile release, modifications are made to the outer layer of the tablet during formulation development to control the lag time prior to release of the drug from the tablet’s inner core (4–5). 

Studies used Penwest’s TIMERx technology (6) (Figure 1) to formulate the tablets, which are comprised of an inner core that contains the drug and an outer layer that has been compression-coated with a hydrophilic matrix of the heteropolysaccharides xanthan and locust bean gum (7). The goal of this work was to formulate compression-coated tablets, taking human circadian rhythms into account, using pulsatile delivery of DS to modulate the drug level for optimal treatment of rheumatoid arthritis (8, 9).   

In this study, a time-dependent compression coated system of DS was developed to target drug release in the colon. If the formulation could be developed so that the required dose of DS could be administered at night, at around 10 PM, it was believed that patients’ morning pain symptoms could be avoided. The drug’s therapeutic effect could also be prolonged by continuously releasing the medication over an extended period of time after the administration of a single dose. 

Factors affecting drug release

Numerous variables are known to affect drug release from time-controlled compression coated tablet formulations. They include the viscosity grade of polymer, the amount of polymer, the drug-polymer ratio, and the nature of the drug used in the tablet system (10, 11). The research summarized in this article set out to investigate the influence of the type and amount of polymer and the viscosity grade of the polymer in the coat on the time-controlled swelling or rupturing of compression coated tablets. 

Materials and methods. DS and hydroxypropylmethylcellulose (HPMC K4M and K15M) were donated by Welable Pharmaceutical Pvt. Ltd. Mehsana. Lactose anhydrous, micro crystalline cellulose (Avicel PH-102), cross-carmellose sodium (Ac-Di-Sol) and magnesium stearate all were purchased from SD Fine Chem Ltd., Mumbai. 

Formulation of core tablets. Direct compression was used to prepare the inner-core tablets of DS. Preliminary trials were run to optimize the core tablet (Table I).

Ingredients

Quantity (%w/w) per tablet

Diclofenac sodium

50

Avicel PH 102

35

Lactose anhydrous

9.5

Cross carmellose sodium 

4

Magnesium stearate

1

Aerosil

0.5

Sunset yellow

q.s.

A powder mixture of the drug, microcrystalline cellulose (Avicel PH-102), lactose anhydrous, cross-carmellose sodium (Ac-Di-Sol), and sunset yellow coloring was dry blended for 30 min in a double-cone blender, after which magnesium stearate and aerosol were added. It is very important to control the mixing rate with magnesium stearate. The blend was further blended for 5 min and compressed into tablets (with an average tablet weight of 200 mg) using a rotary tablet machine equipped with 7-mm diameter flat punches. 

Sufficient pressure was applied to maintain tablet hardness at 6 ± 1 kg/cm². After ejection, the tablets were stored over silica gel in a desiccator for 24 hrs, to allow for elastic recovery and hardening. Prepared tablets were then evaluated for physical properties such as weight and content uniformity, hardness, friability, disintegration time, diameter, and thickness (12,13).

Selection of polymers for compression coating. Studies used two grades of directly compressible polymer hydroxypropylmethylcellulose (HPLC) K4M (3000–5000 mPas for 2% aqueous solution at 20 °C) and HPMC K15M (12,000–21,000 mPas for 2% aqueous solution at 20 °C) for the time-controlled compression-coated formulation of DS. Both grades are known to form gel under aqueous conditions (14–17). 

Preparation of compression coated tablet. A rotary tablet machine was used for compression coating, and the device was maintained at constant pressure, using a two-step direct compression procedure. In the first step, using the weights listed in Table I, the die was filled with half of the weight amount of polymer to make a powder bed, and the core tablet was placed manually in the center of that bed. In the second step, the remaining half of polymer was added to the die above the core tablet. The powder was then compressed under a sufficient compression force, using a 10-mm diameter flat punch to maintain the coated tablet’s hardness at 10 ± 0.50 kg/cm². A standard manual process was used for die filling, core centralization, and machine operation. Data were then analyzed using a 32 full factorial design. In this approach, two factors were evaluated, each at three levels, and experimental trials were performed at all nine possible combinations (Batches F1 to F9). The experimental design with the corresponding formulations is outlined in Table II.

Batch code

Weight of core tablet (mg)^*

Weight ratio of HPMC K15M toHPMC K4M(%) 

Weight of polymer used (mg)

Weight of tablet (after compression coating) mg

P1

200

100:0

150:0

350

P2

200

100:0

200:0

400

P3

200

100:0

250:0

450

F1

200

25:75

37.5:112.5

350

F2

200

25:75

50:150

400

F3

200

25:75

62.5:187.5

450

F4

200

50:50

75:75

350

F5

200

50:50

100:100

400

F6

200

50:50

125:125

450

F7

200

75:25

112.5:37.5

350

F8

200

75:25

150:50

400

F9

200

75:25

187.5:62.5

450

P4

200

0:100

0:150

350

P5

200

0:100

0:200

400

P6

200

0:100

0:250

450

The independent variables were weight ratio of HPMC K15M to HPMC K4M (X1) and amount of polymer (X2). Cumulative percentage drug release after 6 hrs (% Q6), t50- hrs and t75- hours (i.e., time required to release 50% and 75% of the drug respectively) and lag time (hrs) were selected as dependent variables. 

A statistical model (Equation 1) incorporating interactive and polynomial terms was then used to evaluate the responses as: 

     Y = b₀ + b₁X₁+b₂X₂ + b₁₂X₁X₂ + b₁₁X₁₂ + b₂₂X₂₂       [Eq. 1]

Where Y is the dependent variable, b₀ is the arithmetic mean response of the nine runs, and b1 is the estimated coefficient for the factor, X_1. The main effects (X₁ and X₂) represent the average result of changing one factor at a time from its low to its high value. The interaction terms (X₁X₂) show how the response changes when two factors are simultaneously changed. The polynomial terms (X₁₂ and X₂₂) are included to investigate non-linearity (18). 

Statistical analysis. Design Expert (Stat-Ease, Inc. Version 9) software using multiple regression analysis was used to analyze results from testing all factorial design batches. The same software was then used to demonstrate, graphically, the influence of each factor on responses, and to generate response surface plots. Coefficients with a p value less than 0.05 (p < 0.05) were used to look for any significant effects on the model's prediction efficiency for the measured response.

In-vitro drug release of compression coated tablets. In-vitro dissolution studies were then performed on the core tablets and compression coated tablets to verify how the composition of the core and the coat interferes with the drug-release profile. The United States Pharmacopeia–National Formulation (USP–NF) 24 method (basket method, 100 rpm, 37±0.5 °C) was used for dissolution of each of the formulated compression-coated tablets. For the initial two-hour study in 900 mL of simulated gastric fluid, followed by dissolution in simulated intestinal fluid at a pH of 6.8, aliquots of predetermined quantity were collected manually at specific time intervals. Using a UV-visible spectrophotometer at a λ max of 276 nm, they were analyzed for drug content. All dissolution and lag time studies were repeated three times (n = 3). 

Lag time of compression coated tablets. Compression-coated tablets were placed into a USP dissolution paddle apparatus at a rotation speed of 50 rpm with phosphate buffer Indian Pharmacopoeia (IP) pH 6.8, 37±0.5°C and observed visually. The lag time was defined as the time point, when the outer coating ruptured due to swelling/ erosion.

Results and discussion

Pharmacotechnical characteristics of core and compression coated tablets. Core tablets of DS were prepared by direct compression using Avicel PH 102 as a directly compressible excipient, lactose anhydrous as a diluent, and Ac-Di-Sol as a disintegrating agent. The core tablets met all pharmacopoeial requirements in terms of hardness, friability, disintegration time, content uniformity, and weight variation. In 15 minutes of testing, 93% of the drug was released. Upon contact with the dissolution medium, a core tablet began to swell and eventually burst to release the drug. It could be due to porous nature of Ac-Di-Sol. Test results are shown in Table III.

Evaluation parameter

Result

Weight variation (mg) (n=20)

201± 2

Drug content (%) (n=10)

97 ± 2

Diameter (mm) (n=6)

7 ±0.02

Thickness (mm) (n=6)

4.2 ± 0.05

Hardness (kg/cm²) (n=6)

6.1 ± 0.4

Friability (%) (n=30)

0.52±0.2

Disintegration time (min) (n=6)

12±1.0

% drug released at 30 min

93 ±4.0

The compression coated tablets were evaluated for the various pharmacotechnical parameters like weight variation, content uniformity, hardness, friability, thickness etc., and all were found to be satisfactory. 

Results of preliminary batches. The results of in-vitro release profiles of DS from the preliminary batches (prepared with varying amounts of HPMC K15M and HPMC K4M alone in the outer compression coat) showed that the viscosity grade had a marked effect on drug release rate and lag time. Drug release from the formulations with HPMC K4M (P4-P6) could not be controlled up to 24 hrs because it did not swell homogeneously. HPMC K4M's lack of homogeneity was responsible for the more rapid gel layer formation and higher drug release rate. Formulations with HPMC K15M (P1-P3), however, can extensively retard the drug release and form a gel more viscous than that formed with K4M and decrease drug release. Therefore, the study's aim was to adjust the drug release rate by combining low- and high-viscosity grades of HPMC.

Drug release mechanism. Both rupture and erosion mechanisms might be responsible for the time-controlled release mechanism. In-vitro dissolution profiles for all batches are shown in Figure 2. 

Drug release kinetics followed zero order and Higuchi- suggested controlled drug release from prepared compression coated tablets. The results of tests involving experimental batches are given in Table IV. 

Batch

no.

Variables levels in coded form

Cumulative drug release after 6 hrs 

(% Q₆)^*

t₅₀^*

(hrs)

t₇₅^*

(hrs)

Lag time^*(hrs)

X₁

X₂

F1

-1

-1

24.16±1.08

8.3±0.69

11.6±0.85

2±0.43

F2

-1

0

19.28±1.95

9.8±0.88

13.9±0.95

2.5±0.49

F3

-1

+1

15.18±1.00

10.8±0.95

15.1±1.16

3±0.28

F4

0

-1

16.23±1.69

9.9±1.05

14.1±0.37

2.5±0.36

F5

0

0

8.08±1.12

10.7±0.25

14.9±0.80

4±0.25

F6

0

+1

11.26±1.71

11.9±0.50

16.9±0.55

3.5±0.53

F7

+1

-1

10.31±1.85

10.7±0.78

14.9±0.58

3±0.58

F8

+1

0

8.61±1.70

11.4±0.63

15.7±1.21

4±0.47

F9

+1

+1

2.45±0.96

13.1±0.79

18.2±0.70

5±0.78

Translation of coded values in actual values

Variables level

Low (-1)

Medium (0)

High (+1)

X₁: Weight ratio of HPMC K15M to HPMC K4M   (%)

25:75

50:50

75:25

X₂: Amount of polymer (mg)

150

200

250

Results of polynomial equations and surface response plots

Cumulative percentage drug release after six hours (% Q6). The cumulative percentage drug release after 6 hrs is essential because when the dosage is administered at night, say, at 10 pm, symptoms that are experienced in the early morning hours (at approximately 4 am) would be avoided and the drug’s therapeutic effect would be prolonged by continuously releasing the drug over an extended period of time after an administration of single dose. 

Therefore, % Q6 was considered as one of the dependent variables in this research. The value of % Q6 varied from 2.45 to 24.16 for formulated batches and showed a good correlation coefficient (R2= 0.9148). Results from use of the polynomial equation (Equation 2) and surface plot (Figure 3) indicated that both the variables X1 and X2 showed significant effect on % Q6 (p < 0.05). As the weight ratio of HPMC K15M to HPMC K4M increased, the % Q6 value decreased. 

Q6 (%) =+9.96-6.21 X₁ (0.0055)-3.63 X₂(0.0330) +0.28 X₁X₂(0.8505) +2.10 X₁₂(0.3132) +1.90 X₂₂(0.3558)      [Eq. 2]

Time required to release 50% and 75% drug. The t50-hrs and t75-hrs are important measurements that permit the assessment of drug-release profiles, suggesting the amount of drug available at the site. The t50 and t75 varied from 8.3 to 13.1 and 11.6 to 18.2 and showed good correlation coefficient (0.9827 and 0.9682, respectively).

Results of polynomial equations (Equations 3 and 4) and surface plots (Figures 4 and 5), indicated that both the variables X₁ and X₂ showed significant effect on t50 and t75 (p < 0.05). As the weight ratio of HPMC K15M to HPMC K4M increased, the t50 and t75 values increased: 

t50 (hrs) =+10.72+1.05 X₁(0.0005) +1.15 X₂(0.0004)-0.025 X₁X₂(0.8536)-0.14 X₁₂(0.4388) +0.16 X₂₂(0.3983)      [Eq. 3]

t75 (hrs) =+15.03+1.37 X₁(0.0021) +1.60 X₂(0.0011)-0.050 X₁X₂(0.8424)-0.36 X₁₂(0.3116) +0.34 X₂₂(0.3289)      [Eq. 4]

Lag time (in hours). The time at which drug release began in the dissolution medium was taken as an indication for the lag time. The lag time was found to vary from 2–5 hrs and to show a good correlation coefficient (0.9130). 

As the weight ratio of HPMC K15M to HPMC K4M increased, so did the lag time (Figure 6).

This may be due to structural reorganization of hydrophilic HPMC polymer. Increase in the amount and viscosity of HPMC may result in an increase in the gel strength of the polymer and increased mechanical strength of the coating, reducing medium-permeation rate, and resulting in longer lag time.

Result of analysis using a polynomial equation (Equation 5) and surface plot (Figure 7), indicated that both the variables X₁ and X₂ showed significant effect on lag time (p<0.05).

As the weight ratio of HPMC K15M to HPMC K4M increased, the lag time also increased. 

Lag time (hours) =+3.71+0.75X₁(0.0105) +0.67X₂(0.0157)+0.25 X₁ X₂(0.2841)-0.18 X₁₂(0.5371) -0.43 X₂₂(0.1810)       [Eq. 5]

Figure 8 shows overlapping contour plots of all the dependent variables.

The highlighted area in the graph suggests the optimized batch area, which exactly matched the input variables value and output value of Batch F5 (for which the weight ratio of HPMC K15M to HPMC K4M  (%) was 50:50 and the amount of total polymer was 200 mg.

Validation of the statistical model. To validate the statistical model, a checkpoint batch was prepared by taking the level of X1= -0.45 and X2 = 0.45. The values of X₁ and X₂ were then substituted in the equations to obtain theoretical (predicted) values. Comparison of experimental versus predicted values showed that the predicted error varied between -7.14 and +2.77 for prepared formulation. Thus, predicted and experimental values showed reasonably good agreement. Experimental results are given in Table V.

Table V. Experimental responses for the checkpoint batch.

Dependent variables

Observed value

Predicted value

% prediction error

%Q6

8.0

8.3

-3.75

t50( hrs)

5.2

5.1

1.92

t75(hrs)

7.2

7.0

2.77

Lag time (hrs)

1.4

1.5

-7.14

Conclusion

A time-controlled, pulsatile release tablet of DS was successfully developed to enable nocturnal dosing for the treatment of rheumatoid arthritis. When taken at bed time, the dosage form has been designed to release active ingredient in the early hours of morning when patients typically report
experiencing more pain.

In formulating this tablet, lag time was controlled by varying the amount of polymer in the outer compression layer. Drug release, meanwhile, was controlled by varying the nature and viscosity of the grade of HPMC polymer used.   By combining grades of HPMC of different viscosities, and using them in varying amounts, research showed that it is possible to obtain a lag time of 4–6 hrs, and to control release of the drug for up to 24 hrs. Meeting these performance specifications would be necessary for any pulsatile controlled drug delivery system.

References

1. T. Bussemer, et al., Crit. Rev. Ther. Drug Carrier Syst. 18 (5),  433–458 (2001).
2. W. Ritschel and A. Sabouni, J. Contrl.  Rel. 11(5) 97-102 (1990).
3. A. Gazzaniga et al., Eur. J. Biopharm. Vol. 40(4): 246-250 (1994).
4. S. Chaudhari et al., Pharm. Tech., 31(4), 132-144 (2007).
5. S. Lin et al.,  AAPS PharmSciTech 54(5) (2004).
6. G. Patel, et al., Specialized Chronotherapeutic Drug Delivery Systems. Pharmaceutical Reviews 5(1), 2007.
7. A. Baichwal D. Neville, “Advanced drug delivery technology infuses new life into product life-cycles,” Pharmatech Business Brief, 1-5 (2002).
8. I. Kowanko et al., British Journal of Clinical Pharmacology, 11 (5):477-484 (1981).
9. J. Harkness et al., British Medical Journal, 284 (2), 551-554 (1982).
10. B. Janugade et al., International Journal of Chem Tech Research 1 (9): 690-691 (2009).
11. E. Fukui et al., J Control Release, 68(2), 215-223 (2000).
12. L. Lachman et al., The Theory and Practice of Industrial Pharmacy, 4th Ed. (Varghese Publishing House, Bombay, 2004) 243-289, 329-332, 475-501.
13. C. Adeyene, P. Li in Analytical Profiles of Drug Substances, Vol 19, K. Florey, Ed., (Academic Press, New York, 2004). 123-144.
14. S. Lin et al., AAPS PharmSciTech, 54(5)(2004).
15. Dow Chemical Co., Product Information, Methocel Cellulose Ethers: Technical Handbook, Form No. 192-01062-697GW, Midland, MI (2000).
16. M Campos-Aldrete and L. Villafuerte-Robles,  European Journal of Pharmaceutics and Biopharmaceutics, 43(2),173-178 (1997).
17. E. Papadimitriou et al., Int J Pharm. 86(2-3); 131-136 (1992).
18. D. Mayur et al.,  Int J Pharm Investig 2(4), 208–212 (2012). 

Article Details

Pharmaceutical Technology
Vol. 43, No. 11
November 2019
Page: 40-46

Citation

When referring to this article, please cite it as G. Patel, "Optimizing Compression Coating for Pulsatile Release Tablets," Pharmaceutical Technology 43 (11) 2019.