Prediction of Dissolution Profile of Marketed Paracetamol Tablets by

Non-Destructive Near-Infrared Spectroscopy

 

Raagul Seenivasan1,2, Anitha Marimuthu3, Jey Kumar Pachiyappan1,

Gowthamarajan Kuppusamy1, Bhanu Prakash4, Murthannagari Vivek Reddy1,

Vamshi Krishna Tippavajhala2, GNK Ganesh1*

1Department of Pharmaceutics, JSS College of Pharmacy,

JSS Academy of Higher Education and Research, Ooty, The Nilgiris, Tamil Nadu, India.

2Department of Pharmaceutics, Manipal College of Pharmaceutical Sciences,

Manipal Academy of Higher Education, Manipal, Karnataka, India. 576104.

3Department of Pharmacology, JSS College of Pharmacy,

JSS Academy of Higher Education and Research, Ooty, The Nilgiris, Tamil Nadu, India.

4Principal Scientist, SCIENCE4U Analytics and Research Solutions Pvt. Ltd., Bangalore, Karnataka, India.

*Corresponding Author E-mail: gnk@jssuni.edu.in., jeyk984@gmail.com, anitha.marimuthu02@gmail.com. raagul.mcopsmpl2024@learner.manipal.edu

 

ABSTRACT:

Aim and Objective: This study aims to evaluate the efficacy of near-infrared (NIR) spectroscopy in predicting dissolution profiles of six different brands of Paracetamol tablets (500mg) selected based on their prevalence in the local market, considering their higher sales and widespread availability within the geographical area of study. The primary objective of this research work is to establish a robust correlation between the spectral data acquired through NIR spectroscopy and the in-vitro dissolution profiles of these tablets, exploring its potential as a predictive tool. Methodology: The study employs Partial Least Square Regression (PLSR) to obtain distinctive spectral data from Paracetamol tablets using NIR spectroscopy. In vitro dissolution studies are performed to determine the dissolution profiles of the different branded tablets (3 Branded and 3 Generic versions). The collected spectral data is then associated with the dissolution datas using PLSR to create a predictive model. Results and Discussion: The analysis shows a significant relationship between the NIR spectral data and the dissolution values of the Paracetamol tablets. Conclusion: The dissolution analysis revealed strong predictive correlations in branded (A, B, C) and generic (D, E, F) tablet batches. While branded tablets consistently displayed high prediction accuracy, generic versions showcased promising predictive capabilities, warranting minor refinements for improved validation accuracy.

 

KEYWORDS: Dissolution, NIR Spectroscopy, Non-destructive, Multivariant analysis, PLS.

 

 

 

INTRODUCTION: 

Acousto-optic tunable Filter (AOTF) Infrared spectroscopy (IR) is a powerful analytical tool based on the principle of non-destructive spectroscopy. It is the most commonly used vibrational spectroscopic technique for qualitative and quantitative analysis. It is used as a quality control tool in the pharmaceutical industry for testing drug products. It is used as a Process Analytical Technique (PAT) tool, starting from the raw materials to the finished pharmaceutical dosage form to ensure the quality of the drug products. The reason behind the success of this technique is that it is more convenient as they do not require any sample preparation, as they are non-destructive techniques based on molecular vibration. This technique is time-saving and rapid non-destructive testing, and the solvent used in the quality control of the dosage forms is reduced. It is a portable and hand-held device that can be fitted easily to manufacturing equipment and used as a QbD tool1. The Food and Drug Administration (FDA) focuses more on implementing the QbD strategy to improve the quality of the drug product (PAT), which is part of this approach. In the pharmaceutical industry, NIR Spectroscopy is used as either in-line or on-line monitoring of various quality control parameters2, such as moisture content of API and other excipients3, Content uniformity of different pharmaceutical dosage forms4, change in the polymorph of the drug and identification of the adultered drug5. It is used in the real-time monitoring of the manufacturing process and obtaining real-time information about the critical process parameters and a better understanding of such parameters, thus ensuring the quality of the pharmaceutical finished product and reducing batch failure. This technique helps reduce errors and reduces the time of quality checks. It even acts as an important Six Sigma approach where the number of defects per batch can be minimized by continuously monitoring the manufacturing process6. Many process parameters were monitored by this non-destructive approach, such as the blending process of the pharmaceutical ingredients to determine the homogeneity of the composition7-10, the granulation process such as wet granulation, the freeze-drying process, and monitoring the flow of ingredients into the die cavity of the punch. It is also useful in monitoring the dissolution of the drug product11,12. In-vitro dissolution is a part of the Target Product Quality Profile and the QbD approach. It plays an important role in drug product approval and any changes or modifications made in the existing dosage forms16,17. In-vitro dissolution testing is destructive testing, and it is time-consuming and requires more quantity of dissolution media; thus, to overcome this problem, NIR Spectroscopy is used for more effective dissolution testing and reducing the time in the testing of tablets. There are many alternative methods, such as Raman Spectroscopy13. The study involves a large number of data, which are the wavelengths of the scanned tablets, and only useful things are obtained from those wavelengths. For this purpose, the Multivariant Analysis is employed. Multivariant analysis is the statistical approach to examine multiple variables simultaneously. It gives a relationship and interaction between many parameters, such as spectral data, formulation components, and other process parameters, and how they influence the dissolution behavior of the tablet dosage form. By using this approach, it is possible to key patterns and correlations within and among the dataset14. Multivariant Analysis is of many types, one among which is Principal Component Analysis (PCA), and Partial Least Square Regression (PLSR) plays an important role in the prediction of the dissolution profile of pharmaceutical dosage forms. PCA helps to obtain only the required data or information from complex datasets. It transforms the NIR spectral variables or data into a new uncorrelated variable known as Principal components. Another important regression technique is the PLSR technique, which gives the relationship between the spectral data and the dissolution profile of the tablet. It helps find the linear combinations of the original spectral variable associated with the dissolution behavior15.

 

The following are the reasons for predicting the dissolution profile of the commercially available marketed tablets: a) Quality Assurance: It is an important tool in the quality assurance process and ensuring that the marketed products meet the quality and safety where there is continuous monitoring of the drug products. b) Maintain Batch-to-Batch Consistency: Variability among the dissolution of the different batches may affect the therapeutic effects. Thus, this non-destructive technique helps companies identify and sort out such problems and maintain uniformity in the product. c) Expiry date assignment: By predicting the dissolution profile of the drugs, the stability and shelf life of the drug can be determined, allowing for a more precise expiry date assignment. d) Research and development: Helps in the deeper understanding of the dissolution of the drug and formulating the drug molecules accordingly. Thus, reducing the pharmaceutical product recalls. e) Regulatory Aspects: All regulatory bodies focus on the dissolution profiles of the drugs, and by this technique, there is continuous monitoring even after Phase 3 of the clinical trials. f) Validation of the process: By prediction, the manufacturers can validate the production process, thus ensuring the same quality within the batches and bringing robustness.

 

In this research article, the investigation of an alternative non-destructive dissolution method encompasses the evaluation of six commercially available Paracetamol tablets (500mg). Among these, three tablets belong to branded versions, while the remaining three represent the generic counterparts of the branded formulations. The selection of tablets for this research article was based on their prevalence in the local market, considering their higher sales and widespread availability within the geographical area of study. NIR spectroscopy acquired the NIR spectra for each tablet, and the data was preprocessed. The in vitro dissolution of the tablets was obtained by traditional dissolution apparatus USP type 2 Paddle. Using Chemometric software, the two data were correlated, and thus, the prediction of the tablets was done by building a PLS model.

 

MATERIAL AND METHODS:

Materials:

Paracetamol tablets 500mg of six commercially marketed tablets (Brand A, B, C; Generic Brand D, E, F) were purchased from the same Licensed drug store. Instrument used: Labindia dissolution apparatus; Acousto-Optic Tunable Filter (AOTF) Luminar Model 5030, Hand Held Miniature NIR Spectrometer, Version 3; UV Visible Spectrophotometer (Labindia); Software used: Perftest, Brimose Acquire-SNAP32, The Unscrambler, Version: 10.3.0.89 (CAMO Software).

 

Methods:

NIR Spectroscopy:

Acousto-Optic Tunable Filter (AOTF) NIR Spectroscopy (Luminar Model 5030, Hand Held Miniature NIR Spectrometer, Version 3) was used as the non-destructive method for predicting the dissolution of commercially available marketed Paracetamol tablets 500mg of the different brands where Brand A, B, C are the Branded version of Paracetamol tablet and Brand D, E, F are the Generic version of Paracetamol tablets. From each brand, 30 tablets were used in this study. The wavelengths were set ranging from 1100cm-1 to 2100cm-1 and the wavelength Inc was set to 2. These tablets were placed individually in the NIR spectroscopy, and about 30 scans per tablet were made. A total of about 501 background and sample scans were averaged. Spectra were obtained from one side of the tablet because it contains only one API and doesn’t contain any coatings. Software such as Brimose Acquire-SNAP32 was used to record NIR spectra. The NIR Spectra of the tablets are in .dat format, and the spectral datas are obtained from THE Unscrambler, Version: 10.3.0.89 software18.

 

In-vitro Dissolution testing:

The dissolution apparatus (Labindia) was used as a Traditional destructive method for the dissolution testing of the marketed paracetamol tablets. The Traditional method was employed according to the Indian Pharmacopoeia monograph specifications. Dissolution apparatus: USP 2 Paddle type, Rotation Speed: 50 rpm, Dissolution media: Phosphate buffer 5.8 pH, Temperature of the dissolution media: 37±0.5°C, Sampling time: 5, 10, 15, 25, 35, 45, 55; Sample volume: 900ml. 5ml of the sample is withdrawn, and the same amount of the medium is injected. Further dilutions are made accordingly and finally analyzed by UV Visible Spectrophotometer (Labindia) at wavelength 243. Finally, the amount of the drug released was calculated and recorded19.

 

Data Acquisition and Preprocessing of Data:

Acousto-Optic Tunable Filter (AOTF) NIR Spectroscopy was used to scan the tablets on the laboratory scale. The instrument is connected to the computer system where the software BRIMOSE Acquire-SNAP32 was employed to collect the Spectra of the tablets individually and store them in .dat format. These Spectra are converted into numerical datas by the use of The Unscrambler, Version: 10.3.0.89 software. Preprocessing was done to remove the noise or signal that may be caused by factors such as instrument drifts, impurities, and other environmental factors during the testing of the tablet samples. Removing or avoiding these may allow us to use the required spectra and focus on the actual dissolution prediction. Thus, by removing these factors, the precision of the measurement is improved and, thus, accurate quantification. It also provides us with good visualization of the spectral patterns, thus allowing meaningful comparison of the NIR spectral data with the dissolution values. Normalization was done so as to bring all the measurements in a similar scale or range, typically zero and one, thus allowing for an easy and precise comparison. Normalization was carried out in UNSCRAMBLER Software20.

 

Partial Least Square:

After the preprocessing of the datas, the columnset and rowset were defined to carry out the PLS model development. The columnset was divided into Dissolution datas and NIR Spectra data. The rowset was divided between the Training set and the Test set. The Training set consists of the known dissolution datas with NIR Spectral datas. Meanwhile, the test set consists of only the NIR spectral data and unknown dissolution. The PLS model was developed by clicking on the partial Least Square Regression option in the Analysis section of the software. A full cross-validation option was selected, and thus, the PLS model was executed. After Building the PLS model, the prediction was made by clicking the prediction regression option. The option full prediction was chosen, and the below options, such as Inliner limit, Sample Inliner limit, and Identify Outliners option, were selected. In data, Rowset: Test set, Columnset: NIR spectra. In Y Reference, Rowset: Test set, Columnset: Dissolution were chosen, and thus, the predicted dissolution profiles of the tablets appear with the Deviations21.

 

RESULTS AND DISCUSSION:

In-vitro Drug dissolution:

In vitro dissolution was carried out on the six brands of marketed Paracetamol tablets 500mg, and the average percentage drug release of Brand A tables was found to be 71.3%. The average percentage of drug release of Brand B tablets was found to be 86.82%. The average percentage of drug release of Brand C tablets was found to be 78.38%. The average percentage of drug release of Brand D tablets was found to be 73%. The average percentage of drug release of Brand E tablets was found to be 77.1%. The average percentage of drug release of Brand F tablets was found to be 78.4%. The dissolution profile of the marketed Paracetamol tablets are given in Figure 1.

 

Figure 1: In-vitro dissolution of marketed Paracetamol tablets

 

Preprocessing of Data:

The NIR spectra of marketed Paracetamol tablets were normalized due to the noise in the spectral region 1372 cm-1 to ensure consistency of spectral intensity across samples. The spectrum's baseline drift is eliminated, background interference is decreased, and the effective information of the spectra is displayed more clearly. The normalization method provided a uniform baseline, reducing variability unrelated to tablet content. The standard improved data comparability, laying the foundation for further research and observations. NIR spectroscopy with standardized spectra helped to predict tablet properties more accurately. The raw spectral data and preprocessed graphs are given in Figure 2.

 

A

 

B

Figure 2: (A) Raw Data (B) Preprocessed Data

 

Partial least square:

NIR spectroscopy was used to generate scoring plots to test the predictive ability of the model for paracetamol tablet characteristics. The figure 3 shows the relationship between predicted and actual measured tablet asset values. Data points clustered around the diagonal line represent correct forecasts, while scattered points reflect differences between predicted and actual values. A strong diagonal alignment indicates good predictability, but the scattered lines indicate areas of potential model development.  Batch 1: The calibration model yielded an r2 value of 0.992, indicating a strong correlation between predicted and observed dissolution profiles. The validation model exhibited a slightly lower but still significant r2 value of 0.950, confirming the model's reliability. Batch 2: Both calibration and validation models demonstrated high accuracy, with r2 values of 0.995 and 0.981, respectively. These values indicate excellent predictive capabilities for this tablet batch. Batch 3: Similar to Batch 2, Batch 3 displayed robust predictive performance, achieving r2 values of 0.996 (calibration) and 0.987 (validation), signifying strong correlations between predicted and observed dissolution behaviors. Batch 4: While the calibration model showed a good r2 value of 0.989, the validation model reported a slightly lower value of 0.950, suggesting a slight decrease in predictive accuracy. Batch 5: Both the calibration and validation models yielded r2 values of 0.990 and 0.949, respectively, indicating consistent and reliable predictive performance for this tablet batch. Batch 6: This batch demonstrated a decent calibration r2 value of 0.972, but the validation model reported a lower value of 0.880, indicating a potential decrease in predictive accuracy in the validation phase compared to calibration.

 

Figure 3: Score plots of Brand A, B, C, D, E, F

 

Table 1: Prediction of dissolution profile

Tablet

Predicted

Reference

Deviations

Tablet

Predicted

Reference

Deviations

Brand A 1

69.34

70.61

1.26

Brand D 1

64.64

67.02

2.38

Brand A 2

73.08

74.24

1.16

Brand D 2

66.66

68.62

1.96

Brand A 3

74.62

76.43

1.80

Brand D 3

69.89

71.43

1.54

Brand A 4

76.79

78.63

1.84

Brand D 4

71.23

72.85

1.61

Brand A 5

65.98

67.90

1.91

Brand D 5

86.33

88.09

1.75

Brand A 6

68.65

70.75

2.09

Brand D 6

87.25

88.74

1.49

Brand A 7

71.69

72.99

1.30

Brand D 7

68.74

70.38905

1.643306

Brand A 8

73.41

74.68

1.270

Brand D 8

70.30

71.8606

1.554811

Brand A 9

60.19

62.06

1.87

Brand D 9

63.19

64.80155

1.602337

Brand A 10

62.22

64.06

1.83

Brand D 10

64.85

66.35644

1.497015

Brand B 1

85.75

87.11

1.36

Brand E 1

74.96

75.58424

0.6240335

Brand B 2

83.20

84.50

1.30

Brand E 2

76.12

76.73779

0.6132829

Brand B 3

86.82

88.12

1.30

Brand E 3

75.65

76.54886

0.8928545

Brand B 4

85.45

86.86

1.41

Brand E 4

78.12

78.98153

0.8522591

Brand B 5

85.55

87.21

1.65

Brand E 5

70.34

71.09969

0.758625

Brand B 6

83.68

85.41

1.73

Brand E 6

71.34

72.13677

0.7904649

Brand B 7

82.98

84.23

1.24

Brand E 7

80.25

81.27073

1.01707

Brand B 8

81.16

82.36

1.20

Brand E 8

81.68

82.69533

1.005625

Brand B 9

89.36

91.36

1.99

Brand E 9

76.64

77.38503

0.7386748

Brand B 10

89.05

91.39

2.33

Brand E 10

78.29

78.95184

0.6521612

Brand C 1

81.15

81.46

0.30

Brand F 1

80.69

81.73981

1.04547

Brand C 2

73.83

74.14

0.31

Brand F 2

83.13

84.29

1.152

Brand C 3

75.70

76.04

0.34

Brand F 3

75.33

76.70

1.36

Brand C 4

78.29

78.64

0.34

Brand F 4

74.45

75.87

1.42

Brand C 5

71.30

71.62

0.31

Brand F 5

77.43

78.37

0.945

Brand C 6

83.71

84.02

0.31

Brand F 6

79.42

80.54

1.11

Brand C 7

74.74

75.05

0.31

Brand F 7

75.72

76.55

0.83

Brand C 8

82.26

82.58

0.31

Brand F 8

75.86

76.92

1.06

Brand C 9

79.12

79.43

0.31

Brand F 9

75.56

76.44

0.87

Brand C 10

80.87

81.18

0.31

Brand F 10

76.75

77.80

1.05

 

Figure 4: Overall prediction and reference graph

 

Figure 5: Brands A, B, C, D, E, F Prediction Vs Reference

 

CONCLUSION:

The dissolution prediction analysis found strong correlations between predicted and observed dissolving profiles for a large number of tablet batches, proving the efficacy of the dissolution prediction models. While batches A, B, and C of branded tablets consistently showed high prediction accuracy, batches D, E, and F of generic counterparts demonstrated a substantial correlation between predicted and observed dissolution behaviors during calibration phases. Despite significantly lower validation r2 values in batches D and F, these generic versions demonstrated promising predictive skills during calibration. These findings highlight the potential and dependability of the dissolution prediction models for generic tablets, pointing to areas for minor refinement or more research to improve their validation accuracy even further.

 

CONFLICT OF INTEREST:

The authors have no conflicts of interest.

 

ACKNOWLEDGMENT:

The authors would like to thank the Department of Science and Technology – Fund for Improvement of Science and Technology Infrastructure (DST-FIST) and the promotion of University Research and Scientific Excellence (DST-PURSE) for the facilities provided for our department.

 

REFERENCES:

1.      Bhupendra Shrestha, Hema Basnett, P Mohan Raj, Sita Sharan Patel, Mrinmay Das, Neelesh Kumar Verma. Process Analytical Technology: A Quality Assurance Tool. Research J. Pharm. and Tech. 2009; 2(2): 225-227.

2.      Mitesh D Phale. PAT: A New Weapon for Pharmaceutical Industry. Research J. Pharm. and Tech. 2009; 2(4): 611-616.

3.      Mantanus J, et al. Moisture content determination of pharmaceutical pellets by near infrared spectroscopy: Method development and validation. Anal Chim Acta. 2009; 642(1–2): 186–192. DOI: 10.1016/j.aca.2008.12.031.

4.      Li W, et al. Applications of NIR in early stage formulation development. part II. content uniformity evaluation of low dose tablets by principal component analysis. Int J Pharm. 2009; 380(1–2): 49–54. DOI: 10.1016/j.ijpharm.2009.06.032.

5.      Ziιmons E, et al. Determination of binary polymorphic mixtures of fluconazole using near infrared spectroscopy and X-ray powder diffraction: A comparative study based on the pre-validation stage results. J Pharm Biomed Anal. 2011; 55(5): 1208–1212. DOI: 10.1016/j.jpba.2011.02.019.

6.      Chavez PF, et al. Optimization of a pharmaceutical tablet formulation based on a design space approach and using vibrational spectroscopy as PAT Tool. Int J Pharm. 2015; 486(1–2): 13–20. DOI: 10.1016/j.ijpharm.2015.03.025.

7.      Blanco M. Monitoring powder blending in pharmaceutical processes by use of near infrared spectroscopy. Talanta. 2002; 56(1): 203–212. DOI: 10.1016/s0039-9140(01)00559-8.

8.      De Beer TRM, et al. Raman spectroscopy as a process analytical technology (PAT) tool for the in-line monitoring and understanding of a powder blending process. J Pharm Biomed Anal. 2008; 48(3): 772–779. DOI: 10.1016/j.jpba.2008.07.023.

9.      Sulub Y, Konigsberger M, Cheney J. Blend uniformity end-point determination using near-infrared spectroscopy and multivariate calibration. J Pharm Biomed Anal. 2011; 55(3): 429–434. DOI: 10.1016/j.jpba.2011.02.017.

10.   Arti Swami, Shivani Chakankar, Prajakta Chavan, Dnyanesh Limaye, Amol Tagalpallewar. A Critical Review on Recently Used PAT in Pharmaceutical Industry. Research Journal of Pharmacy and Technology. 2023; 16(9): 4436-2. doi: 10.52711/0974-360X.2023.00724.

11.   Gendre C, et al. Development of a process analytical technology (PAT) for in-line monitoring of film thickness and mass of coating materials during a pan coating operation. Eur J Pharm Sci. 2011; 43(4): 244–250. DOI: 10.1016/j.ejps.2011.04.017.

12.   Gendre C, Boiret M, et al. Real-time predictions of drug release and end point detection of a coating operation by in-line near infrared measurements. Int J Pharm. 2011; 421(2): 237–243. DOI: 10.1016/j.ijpharm.2011.09.036.

13.   Mrad MA, et al. Spectroscopy-based partial prediction of in vitro dissolution profile using artificial neural networks. Period Polytech Electrical Eng Computer Sci. 2022; 66(2): 122–131. DOI: 10.3311/ppee.18552.

14.   Zannikos PN, et al. Pharmaceutical Research. 1991; 08(8): 974–978. DOI: 10.1023/a:1015840604423.

15.   Freitas MP, et al. Prediction of drug dissolution profiles from tablets using NIR diffuse reflectance spectroscopy: A rapid and nondestructive method. J Pharm Biomed Anal. 2005; 39(1–2): 17–21. DOI: 10.1016/j.jpba.2005.03.023.

16.   Yu LX. Pharmaceutical quality by design: Product and process development, understanding, and Control. Pharmaceutical Research. 2008; 25(4): 781–791. DOI: 10.1007/s11095-007-9511-1.

17.   Pawar P, et al. Enabling real-time release testing by NIR Prediction of dissolution of tablets made by Continuous Direct Compression (CDC). Int J Pharm. 2016; 512(1): 96–107. DOI: 10.1016/j.ijpharm.2016.08.033.

18.   Baranwal Y, et al. Prediction of dissolution profiles by non-destructive NIR Spectroscopy in bilayer tablets. Int J Pharm. 2019; 565: 419–436. DOI: 10.1016/j.ijpharm.2019.05.022.

19.   K. Soumya, S. Susanna, T.E.G.K. Murthy. In vitro Dissolution Studies on Commercial Brands Containing Immediate Release Pioglitazone and Metformin Hydrochloride Extended Release Tablets. Res. J. Pharm. Dosage Form. and Tech. 2017; 9(1): Jan.-Mar. 15-18.

20.   de Oliveira Neves AC, et al. Dissolution testing of isoniazid, rifampicin, pyrazinamide and Ethambutol tablets using near-infrared spectroscopy (NIRS) and multivariate calibration. J Pharm Biomed Anal. 2012; 57: 115–119. DOI: 10.1016/j.jpba.2011.08.029.

21.   Blanco M, et al. A process analytical technology approach based on near-infrared spectroscopy: Tablet hardness, content uniformity, and dissolution test measurements of intact tablets. J Pharm Sci. 2006; 95(10): 2137–2144. DOI: 10.1002/jps.20653.

 

 

Received on 22.03.2024      Revised on 09.10.2024

Accepted on 13.02.2025      Published on 12.06.2025

Available online from June 14, 2025

Research J. Pharmacy and Technology. 2025;18(6):2582-2588.

DOI: 10.52711/0974-360X.2025.00369

© RJPT All right reserved

 

This work is licensed under a Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International License. Creative Commons License.