Notes
Evaluation of the Effect of Disintegrant Distribution on the Dissolution Behavior of Pharmaceutical Tablets Using Raman Chemical Imaging
2023 年 71 巻 6 号 p. 454-458
詳細
2023 年 71 巻 6 号 p. 454-458
In pharmaceutics, substandard drug manufacturing can sometimes occur. Usually, end-product release tests are conducted to detect defective products, but in many cases, they are not able to identify the root causes of quality defects. In recent years, chemical imaging techniques have been widely used to study quality defects by visualizing the distribution of components in solid dosage forms. However, in most studies, the causes are predicted from images of ingredients, and the impact of each factor is unclear. In this study, we prepared model tablets and intentionally changed only the distribution of disintegrants, and visualized this distribution using the Raman chemical imaging technique to evaluate the effect on the dissolution behavior of the tablets. We found that tablet disintegration occurs completely when the amount of disintegrant is sufficient to disintegrate the tablet and is distributed throughout the tablet, even if the distribution is not uniform. In contrast, if there was a large area where the disintegrant was not present, the tablet did not disintegrate sufficiently. This suggests that it is more important that a sufficient amount of disintegrant is present throughout the tablet rather than the degree of deviation of disintegrant distribution.
Tablets are one of the most commonly used types of pharmaceutical dosage forms. The manufacturing of tablets involves several processes such as blending, granulation, drying, and tableting. In pharmaceutical development, appropriate conditions for these manufacturing processes are determined. However, poor-quality products, the so-called substandard drugs, may occur due to unexpected factors.1–3) Poor-quality pharmaceutical tablets with dissolution delays are sometimes found in the market.4,5) Quality defects attributed to dissolution delay have the potential to affect pharmacokinetics, leading to changes in bioavailability, decreasing the therapeutic efficacy, and increasing the occurrence of side effects. In general, product quality tests are conducted to ensure the quality of pharmaceuticals, and while these tests can detect substandard products, they may not be able to identify the root causes of quality defects; in many cases, this is not a fundamental solution. Thus, investigations have often been performed to determine the cause of the dissolution delay using other techniques. In recent years, chemical imaging techniques6) have been widely used to address this problem. These techniques provide chemical information by continuously acquiring the spectrum of each pixel using analytical techniques such as spectroscopy in two-dimensional or three-dimensional space, and overlaying the obtained spectrum and its coordinate data. Chemical imaging is an effective tool for investigating the causes of quality defects in pharmaceuticals because it provides a more detailed understanding of products and processes by visualizing the distribution of ingredients.5)
Near-IR and Raman spectroscopy are the most common spectroscopic methods used for chemical imaging. They have been widely used in the pharmaceutical field for manufacturing control and evaluation of substandard drugs.7–11) Raman spectroscopy is useful for identifying additives because it detects symmetric stretching vibrations and provides clear spectra with multiple specific peaks. In addition, its high spatial resolution compared with near-IR imaging makes it possible to measure fine particles, such as cornstarch.
The major factors that cause dissolution delay are the physical properties of the active pharmaceutical ingredient and additives, such as disintegrants and lubricants.12,13) The size of the disintegrant has been reported to be one of the most important factors affecting the dissolution of tablets,13) but the effects of the disintegrant distribution on dissolution have yet to be studied. In market surveys, many factors are intricately intertwined in their causes, and it is not possible to reveal the influence of each factor.14) Therefore, it is necessary to evaluate the impact of each factor in advance. In this study, we intentionally changed only the distribution of disintegrants in model tablets and used Raman spectroscopy to evaluate the effects on dissolution behavior.
Two types of disintegrants were prepared in this study: cornstarch (pharmaceutical grade) and crospovidone (Polyplasdone™ XL-10). These were purchased from Japan Corn Starch Co., Ltd. (Tokyo, Japan) and Ashland Inc. (Covington, KY, U.S.A.), respectively. Indomethacin was purchased from Tokyo Chemical Industry Co., Ltd. (Tokyo, Japan) and was used as the active pharmaceutical ingredient. The photograph of these particles are shown in Supplementary Fig. S1. Microcrystalline cellulose (UF-702) was purchased from Asahi Kasei Corp. (Tokyo, Japan) and used as the excipient.
Preparation of the Model TabletsSeven formulations (CS-Comp, CS-60s, CS-10s, CP-Comp, CP-60s, CP-10s, No disintegrants) were used to prepare the model tablets as shown in Table 1. Indomethacin, cornstarch and crospovidone were passed through a 200-mesh sieve (<75 µm) and microcrystalline cellulose was used after passing it through a 100-mesh sieve (<150 µm). To make CS-Comp, CP-Comp and No-disintegrant tablets, all ingredients were blended in an antistatic bag for 5 min. To make CS-60s, CP-60s, CS-10s and CP-10s tablets, indomethacin and microcrystalline cellulose were at first blended in an antistatic bag for 5 min, and the mixture was fed into the die. The disintegrant was then added and blended in the die using a spatula for 60 s with CS-60s and CP-60s, and 10 s with CS-10s and CP-10s. Six tablets with diameters of approximately 8 mm were prepared for each formulation via direct compression at a pressure of 4 MPa for 2 s. Tablets with a single ingredient were prepared following the same procedure to obtain reference spectra for each ingredient.
| Formulation | Cornstarch (mg) | Crospovidone (mg) | Indomethacin (mg) | Microcrystalline cellulose (mg) | Blending time of disintegrants |
|---|---|---|---|---|---|
| CS-Comp | 5 min | ||||
| CS-60s | 60 | 25 | 115 | 60 s | |
| CS-10s | 10 s | ||||
| CP-Comp | 5 min | ||||
| CP-60s | 30 | 25 | 145 | 60 s | |
| CP-10s | 10 s | ||||
| No disintegrants | 25 | 175 | 5 min |
The Raman chemical imaging scheme is presented in Fig. 1. This was performed using a Raman Workstation (Kaisar Optical Systems, Inc., MI, U.S.A.). For each model tablet, the front and back sides were determined, and Raman chemical imaging was performed three areas for each side, six areas in total per tablet. Spectra were obtained at an excitation wavelength of 785 nm, a laser power of 400 mW, an objective lens of 50 ×, a single accumulation, and an acquisition time of 1 s. Raman images were obtained from an area of approximately 1000 × 1000 µm (2601 pixels) with a step size of 20 µm. Reference tablets of the pure ingredients were scanned five times under accumulation and an acquisition time of 10 s to create a spectral library of the ingredients. Data were analyzed in the 1800–200 cm−1 region using Isys chemical imaging software (version 5.0; Malvern Instruments, Ltd., Worcestershire, U.K.). Spectral data were normalized using standard normal variate (SNV) to correct variations in intensity due to physical factors.15) Raman images were generated with the normalized spectral data of the model tablets using Partial Least Squares Discriminant Analysis, a multivariate analysis method based on a spectral library of pure reference ingredients.14) Then, the imaging results from the six areas were concatenated to create a single image. Binary images were created from the disintegrant images after concatenation, and the percentage of area covered per measurement area was calculated. The mean area, standard deviation, and coefficient of variation (standard deviation/mean area) were calculated from the percentage of the area at six measurement points. The coefficient of variation of disintegrants was used as an indicator of variability.
Dissolution tests were performed according to the Japanese Pharmacopeia, using a dissolution tester (DT 126 light; ERWEKA GmbH, Langen, Germany). UV-visible Spectrophotometry was performed using a UV-1280 spectrophotometer (Shimadzu Corp., Kyoto, Japan). The test was performed at 50 rpm according to the Paddle method, using 900 mL of second fluid as the dissolution medium for the dissolution test. After 6, 13, 20, 30, 45, 60, 90, and 120 min from the start of the test, 10 mL of the medium was removed and filtered through a membrane filter (Millex Syringe Filter, Nylon, Non-sterile, 0.45 µm; Merck KGaA, Darmstadt, Germany). This was used as the sample solution. Approximately 28 µg/mL of indomethacin solution was prepared and used as the standard solution. Using the second fluid as the blank, the absorbance of the sample and standard solutions was determined at 223 nm.
The reference Raman spectra of the four ingredients preprocessed by SNV are presented in Fig. 2. As it presents spectral peaks that are specific to each compound, it is easy to discriminate between them and generate chemical images. The Raman chemical imaging results are presented in Fig. 3. Raman images with a uniform distribution of the disintegrants were obtained for CS-Comp and CP-Comp, Raman images showing heterogeneous distribution were obtained for CS-60s and CP-60s, and Raman images showing considerable heterogeneous distribution with regions containing little disintegrant were obtained for CS-10s and CP-10s. The percentage of area covered by each measurement is shown in Table 2. The coefficients of variation for the CS-30% model were 1.9 times for the CS-60s and 4.3 times for the CS-10s compared with the CS-Comp, respectively. The coefficients of variation for the model tablets including 15% crospovidone were 2.8 times for the CP-60s and 8.3 times for CP-10s compared with the CP-Comp, respectively. In both disintegrant-containing models, the coefficient of variation decreased depending on the blending time.
a) CS-Comp.; b) CS-60s; c) CS-10s; d) CP-Comp.; e) CP-60s; f) CP-10s. Red: Disintegrants, Green: Indomethacin, Blue: Microcrystalline Cellulose.
| A) Tablets including 30% cornstarch | |||
|---|---|---|---|
| Area No. | CS-Comp (%) | CS-60s (%) | CS-10s (%) |
| 1 | 23.0 | 48.0 | 85.5 |
| 2 | 44.0 | 30.7 | 20.5 |
| 3 | 33.6 | 36.1 | 33.7 |
| 4 | 16.2 | 35.4 | 0.0 |
| 5 | 34.2 | 6.2 | 0.4 |
| 6 | 42.0 | 6.0 | 0.2 |
| Mean Area | 32.2 | 27.1 | 23.4 |
| Standard Deviation | 9.86 | 15.72 | 30.50 |
| Coefficient of variation | 0.31 | 0.58 | 1.30 |
| B) Tablets including 15% crospovidone | |||
| Area No. | CP-Comp (%) | CP-60s (%) | CP-10s (%) |
| 1 | 20.4 | 49.4 | 72.6 |
| 2 | 25.4 | 29.1 | 2.2 |
| 3 | 21.7 | 40.3 | 6.3 |
| 4 | 27.4 | 7.9 | 0.0 |
| 5 | 24.3 | 16.0 | 0.0 |
| 6 | 11.5 | 5.5 | 0.0 |
| Mean Area | 21.8 | 24.7 | 13.50 |
| Standard Deviation | 5.13 | 16.34 | 26.50 |
| Coefficient of variation | 0.24 | 0.66 | 1.96 |
A: Tablets including 30% Cornstarch, B: Tablets including 15% crospovidone
There were no differences in the dissolution behavior between CS-Comp and CS-60s and between CP-Comp and CP-60s, suggesting that the degree of disintegrant inhomogeneity detected in the Raman images for CS-60s and CP-60s may not affect the dissolution of the indomethacin. In contrast, a dissolution delay was observed in CS-10s and CP-10s, as shown in Fig. 4. Similar tendencies were observed for both model tablets. After 1 h, the model tablets that presented no dissolution retardation (CS-Comp, CS-60s, CP-Comp, and CP-60s) were disintegrated (Fig. 5). Moreover, the amount of residue at the bottom of the vessel was small. In contrast, CS-10s and CP-10s, which showed retardation in the dissolution, had large, aggregated particles and many residues. These dissolution delays were due to poor tablet disintegration.
A: Tablets including 30% cornstarch, B: Tablets including 15% crospovidone.
a) CS-Comp.; b) CS-10s; c) CP-Comp.; d) CP-10s.
Comparing the percent area covered by the disintegrant for each measurement point (Table 2), the CS-10s and CP-10s had areas without the disintegrant, indicating that it was not sufficiently incorporated into the back side of the tablets and, therefore, the disintegration of the tablets was delayed, whereas the CS-60s and CP-60s were inhomogeneous, the area covered by the disintegrant was greater than 5% in any area. This suggests that a sufficient amount of the disintegrant was present throughout the CS-60s and CP-60s tablets. Percolation theory is able to explain that there is a threshold for the amount of disintegrant that affects dissolution.16,17) When the amount of disintegrant is less than the threshold, the network of disintegrant particle in percolation theory is closed and the disintegration of tablet is delayed. Therefore, it is considered that the dissolution of the CS-10s and CP-10s, in which the disintegrant is partially not distributed, were delayed. It is important not to generate areas with no disintegrant in the tablet, and it is possible to determine that the quality of tablets with areas without disintegrant distribution is poor. However, it is difficult to quantitatively evaluate dissolution only by two-dimensional imaging, because the distribution of the disintegrant in the tablet is three-dimensional. Therefore, evaluation with three-dimensional imaging technology is an issue in the future.
We investigated the effect of the distribution of disintegrants in model tablets on the dissolution behavior using Raman chemical imaging. We found that when the disintegrants were distributed throughout the tablets, disintegration occurred completely, whereas if the disintegrants were not widely distributed, the tablets would not disintegrate sufficiently. Our results show the importance of having a sufficient amount of disintegrant throughout the solid dosage forms for effective dissolution, rather than the degree of deviation of disintegrant distribution.
This research was supported by the Japan Agency for Medical Research and Development (Grant number 22ak0101190).
The authors declare no conflict of interest.
This article contains supplementary materials.
