OR WAIT 15 SECS
New technologies, such as NIR and Raman, enable online measurements of blending and content uniformity in the production of solid dosage forms.
Advanced analytical instruments, such as near infrared (NIR) and Raman spectroscopy, have revolutionized solid dosage manufacturing by enabling real-time measurements of critical quality attributes such as content uniformity. Content uniformity testing is an important assessment for oral solid dosage (OSD) forms. “It is one of two allowable tests to assess the uniformity of dosage units (UDU), the other being weight variance,” says Darren Andrews, Pharma Business Manager, Raman Spectroscopy, Cobalt Light Systems, part of Agilent Technologies. “UDU is a test of the variance of the active ingredient over the batch manufacturing process.”
Content uniformity is an important quality measure of the final solid dosage product, notes Ian Robertson, spectroscopy applications specialist at PerkinElmer. It ensures that a consistent dose of the API is maintained between batches so that the patient receives the correct dose. Robertson points out that content uniformity is particularly important where tablet splitting is used. “The active ingredient needs to be evenly distributed throughout the tablet to ensure that if the tablet is split in half, each half of the tablet has an equal dose,” he says.
Content uniformity testing sets a limit on the variance of API within each tablet or capsule. “The intention is to make sure that OSDs do not have an abnormally low or high amount of API, which may arise because powder is blended and processed before being tableted or encapsulated,” Andrews says. “Blending will cause natural variations in the level of a dosage amount.”
Blending a large amount of bulk powder such that each small amount has a total API content within a few percent of the target concentration is a significant challenge, observes Andrews. “A tablet is often tens or hundreds of milligrams in weight and is pressed from a small amount of a powder blend that can weigh more than one tonne. Although the processes that make the blend are designed to minimize variance, it can’t be eliminated completely,” he says. “The bulk material may change as it is transported, which can change the composition of the blend in the container as a function of powder depth. Sometimes, the beginning and end of a batch run can be out of specification because of this composition change and are, therefore, discarded.”
“The aim of blending is to produce a homogenous blend of a sample that accurately represents the ratios of the components within it,” Robertson underlines. “Proper blending is essential to ensure consistent dosage and performance of solid dosage forms.” He highlights that each different blend will require unique blending conditions.
“Non-homogeneities can be caused by many factors and can be related to the materials to be blended, the type of machinery used, and the processing parameters used for the blending,” Robertson explains. “The most common problems are insufficient blending time and overblending. For any given blend, under any given processing conditions, there will be an optimum blend time. With insufficient blending time, the materials do not have the possibility to achieve a homogenous mix. The obvious thought would be to just continue blending for a much longer time. However, this would greatly increase manufacturing costs and can also lead to overblending, which can cause demixing and segregation of the mixture.”
According to Robertson, the type of blender needs to be appropriate for the materials being blended and whether it will be a batch or continuous process. “The size and shape of the blender must be appropriate for the blend being processed,” he says, emphasizing that the processing parameters of the blending process play a crucial role in avoiding non-homogeneity. “Non-homogeneity of the blend will lead to inconsistent dosage in the final product and could lead to product rejection, which will result in significant financial losses. Re-blending of a non-homogenous product is generally unfeasible, and the product needs to be scrapped.” Robertson recommends testing the blend during or after blending to prevent inhomogenous blends from being processed further.
Once a suitable type of blender has been selected, it is important that the formulation and processing variables are optimized for that particular blend. “The major processing variables are mixing time, speed of mixing, and powder fill volume,” says Robertson. “Each of these parameters will, individually, have an effect on the blending efficiency.”
“In rotational blenders, the mixing time is the product of the number of rotations and the speed of rotation of the blender,” Robertson explains. “The speed of mixing (or speed of rotation) will affect the quality of the blend. For free-flowing materials, the speed can be quite high and the materials will efficiently mix. However, for cohesive materials, the speed needs to be considerably slower to allow the ingredients to mix well.” He notes that it is also important to determine the optimum powder fill volume within the blender. “Too large a fill will prevent mixing during the blending process,” he says.
Robertson stresses that during the blending process, it is necessary to determine the blend homogeneity to determine completeness of blending. “This has traditionally required a sampling strategy to determine where and how often samples are taken for measurement,” he says. “Sampling was primarily performed using sample thieves, which have their own shortcomings.” Thief sampling is not instantaneous, produces little data, and has many sources of error and bias.
Guidance such as United States Pharmacopeia (USP) <905> (1), European Pharmocopoeia (PhEur) 2.9.40 (2), and other pharmacopeia chapters have set out the requirements for UDU by content uniformity testing. Generally, only 10 tablets from a batch have to be tested, Andrews notes. “These samples are assayed individually and a calculation using the individual values is made to find an acceptance value,” he says. “This method spreads the allowed variance over the test samples, and depending on the variance of the samples, an individual may have >10% deviation from the nominal value and the batch could still pass the test.”
According to Andrews, using 10 samples to describe the batch variance is a limited number to assess the UDU of a large batch. He explains that if the test fails with n=10, taking an additional 20 samples to make n=30 is allowed, and the content uniformity test may pass as a result. “Taking more samples from across the batch routinely would give greater confidence in the batch process,” he says. “Some regulators are expanding the options for testing a larger number of samples, which becomes more feasible using spectroscopic technologies. PhEur 2.9.47 (3), for example, provides options for testing 100–10,000 individual samples.
Robertson points out that the biggest challenge in content uniformity testing for solid dosage forms is the sampling strategy, “namely, how do you ensure you are getting representative samples that reflect the whole batch?” They note that the results from USP <905> lack confidence because the method does not use a statistical sampling plan, and therefore, provides limited assurance that the batch meets specifications and quality control criteria. In fact, FDA no longer supports the use of USP <905> for product release testing. FDA also withdrew, in August 2013, its guidance document, Powder Blends and Finished Dosage Units-Stratified In-Process Dosage Unit Sampling and Assessment (4), because it no longer reflected the views of the agency (5). In response, a group of individuals from industry, academia, and FDA formed the Blend and Content Uniformity (BUCU) team, within the International Society for Pharmaceutical Engineering (ISPE) organization, to come up with alternative approaches and best practices for assessing blend and content uniformity. The team has developed modifications to the withdrawn FDA draft stratified sampling guidance and proposed sampling plans based on the application of two ASTM methods: E2709 (6) and E2810 (7).
The US Pharmacopeia Convention has invited the BUCU team to advise on the preparation of the new chapter USP <1905> on sampling considerations for batch release (8). ISPE has launched a section on its website to make available to the industry the BUCU team’s output.
For OSDs, there needs to be a suitable sampling strategy for content uniformity testing, Robertson highlights. “The most common sampling plans suggested by the BUCU team are: simple random sampling, stratified sampling (partitioning the batch into sections, then sampling from each section), and systematic sampling (at regular intervals throughout the batch),” he says.
The most common method for assessing content uniformity in OSDs is high-performance liquid chromatography (HPLC), observes Andrews. “HPLC has the advantages of flexibility, sensitivity, and ubiquity, and many analytical chemists are trained to use HPLC methods,” he says. “The disadvantages, however, are the time and resources needed to prepare samples.”
“In blending operations, the sample is removed using a sample thief and analyzed by HPLC using a calibration curve for the active ingredient(s),” says Robertson. He explains that the method involves dissolving the tablet in a suitable solvent and analyzing the amount by HPLC against a calibration. Although accurate and repeatable, the disadvantages with HPLC are that measurements are offline, samples have to be removed from the blending process, it is destructive for tablets, and analysis can take several minutes per sample, Robertson highlights.
Ultraviolet/visible (UV/Vis) spectrophotometry is another technique that can be used as an assay for content uniformity measurements. “The absorption at a known, given wavelength for a material will be directly related to the concentration of the material present,” Robertson says. “Using this correlation requires the material being analyzed to have a suitable absorption within this spectral range and for the other materials present not to have interfering absorptions, although spectral overlap can generally be accounted for by spectral deconvolution or appropriate curve-fitting quantitative methods.” He, however, points out that like HPLC, UV/Vis spectrophotometry is an offline and destructive method.
Over the past decade, there has been an increasing emphasis on quality by design (QbD) and the use of process analytical technology (PAT) to monitor and control pharmaceutical manufacturing processes. “FDA encourages the use of new technologies and PAT techniques, such as NIR spectroscopy, for content uniformity analysis,” says Robertson. “NIR spectroscopy can be used throughout the manufacturing process in online, at-line, and offline modes of operation.” He explains that one such application involves the online monitoring of the blending process with small NIR instrumentation fitted directly onto the blender, or NIR instrumentation connected to the blender using fiber optics probes. “This set-up allows for real-time measurement of the blending process without the requirement to remove samples for measurement, which is a significant advantage over other techniques, such as HPLC,” Robertson says. “Different mathematical approaches can be applied to the spectral analysis to determine when the blend is optimized, without the requirement for quantitative calibrations.”
Robertson adds that NIR spectroscopy can also be applied to the measurement of content uniformity of solid dosage tablets. “The measurement can be performed non-destructively directly on the tablet using reflectance or transmission after suitable quantitative calibrations, typically against reference values from HPLC measurements,” he says. “No sample preparation is required, and the measurement does not require solvents or other hazardous chemicals.”
As is the case for NIR, Raman instrumentation enables non-destructive, online measurements of the pharmaceutical process, including blending and content uniformity measurements of tablets. “Raman spectroscopy can be used as a PAT tool for content uniformity applications,” says Karen Esmonde-White, senior marcom specialist at Kaiser Optical Systems. “The sampling flexibility of Raman lends itself to an offline measurement using a Raman microscope or sampling compartment and an in-process measurement using sampling probes.” She points out that in-process Raman measurements are beneficial because they provide real-time process understanding without needing to collect samples for offline measurements.
“The real-time process understanding provided by Raman enables in-process corrections, improves efficiency, ensures quality, and can form the basis of automated feedback control,” Esmonde-White explains. “The specificity of Raman spectra, sampling flexibility, and compatibility with aqueous media are advantages over other spectroscopic techniques. Raman spectroscopy allows quantification of the API within the excipient matrix, and customers have reported API quantification to sub % levels using Raman.”
Robertson concurs that Raman has the advantage of being extremely specific in terms of the spectral profiles of different materials present in formulations, whereas NIR spectra typically contain significant spectral overlap of the ingredients. He highlights that Raman, however, does have the disadvantage of being a weak scattering technique and is not suitable in cases where the samples fluoresce.
According to Robertson, a limitation of Raman spectroscopy for content uniformity measurements has been the limited spot size of the laser, with only a small fraction of the complete tablet being analyzed. “However, recent advances using raster scanning of the laser allow the entire tablet to be sampled,” he say. “Further advances have been achieved using Transmission Raman Spectroscopy (TRS), in which the Raman signal transmitted through the whole tablet is measured, thus measuring a much larger sample volume.”
TRS is an alternative spectroscopic technique to quantify ingredients in a dosage form that has been available for approximately eight years, according to Andrews. “It is fast and flexible and can analyze hundreds of samples per hour,” he says. “TRS works with a wide variety of tablets and capsules and is suitable for high-volume testing, with significant impact on resource savings.”
“It is Raman spectroscopy and so is complementary to infrared spectroscopy; both have information-rich spectra,” Andrews explains. He adds that by testing intact tablets, the solid-state form (e.g., polymorph type) can also be quantified.
TRS is not an absorption technique, Andrews points out, hence, it is less sensitive to sample thickness than NIR. “A single measurement contains the data needed for analysis, allowing multiple APIs, excipient concentrations, and more to be determined with no additional effort.” TRS is currently used to release tablets and capsules with regulatory-approved tests for content uniformity, assay, and drug product identification.
Having in place control procedures to monitor the blending process is key to ensuring consistent and efficient production of OSDs. PAT tools, such as NIR and Raman, provide an alternative to sample thieves by enabling online analysis of content uniformity.
1. USP, USP General Chapter <905> Uniformity of Dosage Units (US Pharmacopeial Convention, Rockville, MD).
2. EDQM, PhEur General Chapter 2.9.40 Uniformity of Dosage Units (EDQM, Strasbourg, France).
3. EDQM, PhEur Chapter 2.9.47 Uniformity of Dosage Units Using Large Sample Sizes (EDQM, Strasbourg, France).
4. FDA, Guidance for Industry, Powder Blends and Finished Dosage Units-Stratified In-Process Dosage Unit Sampling and Assessment (Rockville, MD, October 2003).
5. J Bergum et al., ISPE Pharmaceutical Engineering 34 (2) (March/April 2014).
6. ASTM Standard Number E2709: Standard Practice for Demonstrating Capability to Comply with an Acceptance Procedure, September 2009.
7. ASTM Standard Number E2810: Standard Practice for Demonstrating Capability to Comply with the Test for Uniformity of Dosage Units, October 2011.
8. ISPE, “ISPE Group Working on Blend and Content Uniformity Launches Website,” Press Release, Sep. 16, 2014.
Vol. 42, No. 2
When referring to this article, please cite it as A. Siew, “Analyzing Content Uniformity,” Pharmaceutical Technology 42 (2) 2018.