|
Research Studies - Renhite Near-Infrared Spectroscopy (NIRS) Quality Assurance Analysis for LaSenti-CF
Abstract
The purpose of the study is to check the quality control, quality assurance, blend uniformity and final dosage form of LaSenti-CF. This study will be performed on an ongoing basis to make sure all samples provided to the LaSenti-CF customer base is similar against the reference standard and methods of analysis using Near-Infrared Spectroscopy (NIRS).
Introduction
In any development or manufacturing environment the product must be subjected to a series of tests to ensure its quality and consistency. In the pharmaceutical and biotechnology industries there is a widespread use of analytical chemistry methods and technologies to analyze and test these quality control measures. This study will be used to perform a product quality control, quality assurance, blend uniformity and final dosage form analysis. This analysis is performed as a post-manufacturing measure and not as a process analytical technology (PAT).
Method
The samples are analyzed using a technology called Near-Infrared Spectroscopy. A reference standard of active pharmaceutical ingredients (API) and excipients for samples analyzed are created. The reference standard of high quality samples are provided to develop a product database that will be used to method development and validation. Five (5) samples are triturated and stored in glass vials to be scanned.
A library for quantitative analysis with a minimum of three sample spectra and a valid reference standard is developed. The purpose of this library is to validate the reference spectra stored in the product database project.
Samples from the product database are created into a library. The sample is selected and a math treatment is applied. We performed a Mahalanobis Distance in Principal Component Space taken out to the second derivative spanning 400 to 2500 nanometers for wavelength. In statistics, Mahalanobis distance is based on correlations between variables by which different patterns can be identified and analyzed. It is a useful way of determining similarity of an unknown sample set to a known one. The detector use type is reflectance. Our outlier threshold is set at 0.95. This level of correlation is kept to be set as an acceptance level for passing or failing.
Once the samples from the product database have been added to a library and applied to math treatments we bring it to the next phase of identifying the method using correlation in wavelength space to the second derivative. This phase will also use the same parameters spanning 400 to 2500 nanometers for wavelength and a threshold of 0.95.
After the library has been validated the instrument is selected. For this study we used a FOSS NIRSystems XDS Near-Infrared Rapid Content Analyzer with a spot seizer to centralize the vials analyzed on the window of the content analyzer.
Routine analysis is ran and the machine takes readings from the sample being analyzed and correlates that sample against the validated library. The closer the analysis is the 1.00 the greater the correlation between the analyzed sample and the validated library that was built with reference standards.
Results
The results show below are the readings and spectra taken from the NIR machine. The samples analyzed for this range from A-1 to A-3 and B-1 to B-3. The red colored line (spectra) is the validated library and the blue line represents the analyzed sample.
Conclusions
The level of acceptance for pass or fail is 0.990 for correlation. This level of acceptance is high but all of LaSenti-CF's samples have met and exceeded that mark. Therefore, with the 6 samples analyzed all samples have passed analysis.
|
