Opportunities and Pitfalls in the Analysis of Subvisible Particles during Biologics Product Development and Quality Control

Introduction

Subvisible particulate matter testing in injectable drugs has been required for many parenteral drugs since USP <788> was first implemented in the early nineteen eighties.¹ The acceptance criteria has not changed through the years with a focus on particle counts at 10 and 25 μm per container for small volume parenterals (SVP) and large volume parenterals (LVP), see Figure 1.

Now that protein based therapeutics are more common, USP<787> was written to accommodate the differences in the products and sample volumes. Other trends in both industry and regulatory oversight have been to better understand and quantify protein aggregation and associated immunogenicity.^2,3

The relationship between interest in quantifying protein aggregation and enhancing particle counting measurement techniques at smaller sizes (below 2 um and submicron range) is driven by multiple influences including:

• Protein products contain a wide size range of aggregates, nm – μm and it is highly useful to characterize and better understand what sub-population is more relevant to drug safety
• Subvisible Particles are critical species on the Protein Aggregation Pathway
• Subvisible particle counts provides a sensitive indication of protein aggregation
• Formation of subvisible particles (nano& micro) is an early step on aggregation pathway and precursor to visible particulate formation
• Even trace levels of particles can impact subsequent stability of protein solutions; aggregates beget aggregates

Other factors contributing to the interest in extending the range of analytical techniques include the need to better understand the relationship between the stress induced by manufacturing conditions and protein stability. These stress conditions include:

• Physical/chemical Stresses: pH, ionic strength, temperature, chemical modification, light, agitation, mechanic shock, freezethaw, etc.
• Air/Solid-Liquid Interfaces: Protein solution contact with pumps, pipes, vessels, filters, chromatography columns, etc.
• Existence of foreign Particles that can destabilize the protein drug: Stainless steel, glass, plastic, rubber, silicone oil, etc.⁴

New analytical techniques are being adapted to provide additional insight into predicting and quantifying protein aggregation. For example, Flow imaging microscopy (FIM) is used to obtain size and morphology information along with images that can help identify particulate species such as oil droplets vs. aggregated proteins. It is expected to see the FIM technique appear in the USP tests sometime soon.

The FDA currently requests data for subvisible particles between 2 – 10 μm from studies using quantitative methods and has suggested that information on the relationship between subvisible particle content below and above 2 μm may be informative for setting limits.⁵ The current focus on a lower limit of 2 μm is connected to the historic lower limit of light obscuration sensors, which is based on the principle of optical extinction. But combination extinction/scattering sensors have been on the market since the mid nineteen nineties; and this means that the ability to cover the dynamic range of 0.5 – 400 μm is well established. Working with low volume protein based therapeutic injections can be challenging even with the lower volume requirements for USP 787.⁶ The article will focus on how one can take advantage of a well-established and robust technology for measurement of submicron and micron-sized particles while avoiding potential pitfalls.

The Challenges of Lower Sample Volumes

The USP <788> test procedure requires 20 mL sample for each test; four measurements of 5 mL each. The first result is discarded and the average of the next three results are reported at 10 and 25 μm.¹ The approach of discarding a first test and taking the average of multiple measurements is accepted good practice in the field of particle characterization. But the desired sample volumes for analytical techniques for protein based therapeutics are smaller than historic small molecule products for reasons of both product value and available quantity. For these reasons, the USP <787> test notes lower volumes possible, on the scale of 0.2 – 5 mL.⁷ Sample volumes of 200 μL are possible with optical particle counters, but this requires some changes in how the measurements are made.

Determining the minimum sample volume for a given analytical technique should involve a careful study of a known sample and results should be carefully reviewed for both accuracy and repeatability. Both the tare and sample volume should be considered and sample tubing should be chosen taking into account both results generated and practical sample throughput considerations.

To better understand how to develop the optimal test parameters, a study was performed to determine the realistic lower sample volumes required for an optical particle test. The sample used to perform this study was a 15 μm particle count standard from Micro Measurement Labs, Inc., lot #NK20C. The reference count value for this standard is 3,118 – 4,218 particles/mL. All measurements were made on the Entegris AccuSizer SIS system equipped with the LE400 sensor, calibrated and used at a flow rate of 15 mL/min. A 1 mL syringe was installed onto the SIS sampler. The measurement procedure used is described below:

1. Flushes of 0.5 mL were performed before and after sampling (an air gap took place after the ‘before’ flush, but before the sampling).
2. Fresh 900 μL aliquots were used for each sample, regardless of sampling required.
3. An air gap of 0.05 mL was used in each run prior to sampling.
4. A tare volume of 0.15 mL was used for each measurement. Measurements were performed at the following sample volumes: 650, 550, 450, 350, 250, 150, and 50 μL. All measurements were performed in triplicate.

The results from this study are shown in Figure 2.

Another experimental parameter worth consideration when minimizing sample volume is mixing. Mechanical stir bars can’t fit into very small sample containers or well plates when using automated techniques. The older approach of mixing by hand during the measurement is cumbersome and cannot adapt well to automation. One new approach is push/pull mixing using the syringe sampler itself. The sample is pulled up through the sensor and pushed back into the container multiple times prior to the actual analysis.

Subscribe to our e-Newsletters
Stay up to date with the latest news, articles, and events. Plus, get special offers
from American Pharmaceutical Review – all delivered right to your inbox! Sign up now!

When working with protein based therapeutics it is worth investigating if this push/pull mixing introduces enough shear and/or interfacial stress on the sample to induce aggregation. Parameters to study include the inner diameter (ID) of the sample tube/needle and the number of times the sample is transported. If the sample undergoes three push/pull mixes and then 4 measurements the sample makes 10 trips [(3x2) + 4] through sample tube/needle. Previous studies have reported on the effect of interfacial stress on protein aggregation (8). When done properly in a systematic way, this measurement technique can essentially serve as a method to study small scale interfacial stress impact on the protein. For this study, a short investigation was carried out before performing the studies shown in this paper and a larger ID needle was chosen to minimize the interfacial stress experienced by the protein samples.

Experimental Method

Materials: The protein used was NIST reference material 8671 (NISTmAb), humanized IgG1κ monoclonal antibody lot number 14HBD-002, expiration date April 2021, concentration 10 mg/mL. Two vials of the same lot number were used (“older” and “new”). NISTmAb is a homodimer that has undergone biopharmaceutical industry standard upstream and downstream purification to remove process related impurities with a molecular weight of approximately 150 kDa. The intensity mean diameter was analyzed using the Entegris, Inc. Nicomp DLS system with a 35 mW laser at 658 nm wavelength and high gain avalanche photo diode detector at 90°. The information value for observed average hydrodynamic diameter by dynamic light scattering (DLS) is 9.96 nm.

The protein aggregates were measured using the Entegris, Inc. AccuSizer A2000 MPA microplate analyzer with the model LE400 sensor, dynamic range 0.5 – 400 μm. This instrument is based on the principle of single particle optical sizing (SPOS), an advancement on the older technique of light obscuration. The intensity mean diameter was analyzed by DLS at a concentration of 1 mg/mL, temperature of 23 C, measurement duration of 7 minutes at concentration of 1 mg/mL. The reported intensity mean was 10.2 nm.

Measurement protocol: A well in the microplate was filled with 1.5 mL of the diluted sample. Mixing was performed using the push/pull technique prior to the first analysis. The tare volume was 100 μL and sample volume of 250 μL. Four analysis were performed, the first discarded and then the last three were averaged. The sample was pushed to waste after every measurement, although sample preservation is an option.

Results

Aggregation Study # 1; New vs. older NIST 8671 sample

A plot of particle count/mL vs particle size for the new vs. older sample is shown in Figure 3. The higher count in the 1 micron and below range is an indication that the older sample had experienced a higher degree of aggregation, presumably due to the additional freeze, thaw cycle and possible aging.

Study 2: Diluent selection; PBS vs. DI water

It is well documented that proteins are typically more stable in buffers such as phosphate buffer saline (PBS) than in distilled water (DI). A short study was performed to investigate the effect of diluting the NIST 8671 mAb in filtered PBS vs. filtered DI water. First the diluents were analyzed to establish an acceptable background. Then 25 μL of NIST 8671 was diluted into 9.9 mL of filtered PBS and DI water. The results are shown in Figure 4. While the DI water reported a lower background particle count than the PBS, the diluted protein counts were lower in the PBS than in the DI water. This indicates that dilution in DI water probably caused a greater degree of aggregation.

Study 3: Before and after heat stress at 60°C at three concentrations.

Sample preparation: Exposing the protein sample to elevated temperatures can cause unfolding and aggregation. Samples were prepared at three concentrations; 20, 30, and 40 μL diluted into 9.9 mL filtered DI water. The samples were analyzed immediately after preparation, and then after four, and twenty hours heat exposure at 60°C.

The same measurement protocol as described above was followed for all measurements.

Results: Figures 4-6 show concentration in particle count/mL vs. size for the unstressed (T0), after fours hours (T4), and after 20 hours (T20) heat stress at 60°C.

The increase in particle count after exposure to heat stress tracks the expected aggregation behavior. The after heat stress concentrations for the 40 μL exceed the sensor concentration limit at 0.5 μm, but that does not change the study conclusion. As shown in figures 1 to 3, all samples would pass the USP<787> evaluation criteria at the 10 μm and 25 μm size range. However, data below 2 um provides much greater insight into the changes that are occurring as the protein is exposed to longer periods of thermal stress. With further method development and optimization, the wider dynamic range of the combined light obscuration and light scattering detector may enable the application of SPOS technology in both formulation and process development.

Conclusions

Several learning points were discovered over the course of these studies. First, additional care must be taken when performing low volume optical particle counting measurements. The choice of sample tube or needle should be investigated to avoid shear or pumping induced aggregation during the analysis. Push/pull mixing appears to be a valid alternative to assure a well-mixed sample during the analyses. Finally, this study shows that, for product development purposes, the most useful data is generated from particles smaller than 2 μm; and reporting and studying submicron particles can provide additional insight into protein aggregation mechanisms.

References

1. United States Pharmacopeia, <788>: Particulate Matter in Injections
2. Singh SK, Afonina N, Awwad M, et al. An industry perspective on the monitoring of subvisible particles as a quality attribute for protein therapeutics. J Pharm Sci. 2010;99(8):3302-3321.
3. Food and Drug Administration (US) Guidance for Industry: Immunogenicity Assessment for Therapeutic Protein Products. August 2014
4. Singh SK, Toler MR. Monitoring of subvisible particles in therapeutic proteins. Methods Mol Biol. 2012;899:379-401.