Quantification and Characterization of Micrometer and Submicrometer

Jul 13, 2012 - Late Stage Pharmaceutical Development, Genentech, 1 DNA Way, South San Francisco, California 94080, United States. •S Supporting ...
0 downloads 0 Views 1MB Size
Article pubs.acs.org/ac

Quantification and Characterization of Micrometer and Submicrometer Subvisible Particles in Protein Therapeutics by Use of a Suspended Microchannel Resonator Ankit R. Patel,* Doris Lau, and Jun Liu Late Stage Pharmaceutical Development, Genentech, 1 DNA Way, South San Francisco, California 94080, United States S Supporting Information *

ABSTRACT: The ability to characterize micrometer and submicrometer particles in solution is of fundamental importance to understanding the relationship between protein particles in biotherapeutics and concerns raised regarding immunogenicity. While a number of characterization methods are available for analyzing subvisible particle content in protein pharmaceuticals, counting and characterizing particles within the entire subvisible size range remains a significant challenge due to the properties of the proteinaceous particles themselves and to the limitations of the available techniques. Additionally, as silicone oil-lubricated prefilled syringes become a favored primary packaging for biotherapeutic products, proteinaceous subvisible particle characterization is further complicated by the presence of silicone oil droplets in solution. Here, we critically evaluate and apply a novel method for particle characterization that relies on differences in particle buoyant mass to characterize particle content in the range of ca. 0.5−5 μm. A model particle system was specifically designed to evaluate the ability of the suspended microchannel resonator (SMR) to distinguish between buoyant particles (e.g., silicone oil) and dense particles (e.g., protein particles) in aqueous solution. In addition, this emerging technique was successfully applied to high-concentration monoclonal antibody solutions stored in prefilled syringes in stressed stability studies. It is shown that the SMR system can potentially distinguish between silicone oil droplets and protein particles in a size range that is challenging for many subvisible particle characterization methods. Limitations of the SMR method are also discussed.

C

over the entire particle size range still remains a significant challenge. This is particularly the case for particles smaller than 5 μm due to the trace levels at which they are usually found in biotherapeutics, the physical properties of the particles themselves, and the limitations of the techniques available to characterize particle content. Recently, several advances have been made regarding the characterization of subvisible particles beyond the compendial methods specified in the pharmacopeia.12 Specifically, the adoption of microflow imaging has been advocated as a potential alternative method to distinguish between irregularly shaped particles and spherical air bubbles and silicone oil droplets.13−15 In addition, the use of electrical sensing zone measurements (Coulter counter) is being revisited,16−18 and advances in single-particle tracking have also been employed for subvisible particle size analysis.19 Methods relying on differences in refractive indices or optical contrast between the particle and solution (e.g., light obscuration, flow microscopy, etc.) are limited by the translucent nature of protein particles and have a tendency to

haracterization of subvisible particles in biotherapeutics has been the subject of intense interest over the last several years1−3 due to concerns regarding their potential for triggering immunogenic responses to protein therapeutics.1,4,5 As a result, regulatory health authorities are routinely requesting new drug sponsors to provide additional characterization of the amount, size, and identity of subvisible particles smaller than 10 μm when reviewing regulatory filings for clinical trials and product licensure. While the nature of the relationship between subvisible particles and immunogenicity or product safety remains under debate,4,6−9 accurate characterization of the subvisible particle content in parenteral formulations of protein therapeutics is critical to assessing this relationship. In literature, the terms “subvisible particles”, “submicrometer particles”, and “soluble aggregates” have been used to describe particles in different size ranges below the visible range,4,10 and a continuum of aggregate and particle sizes likely exists in biotherapeutic formulations.11 For the purpose of this paper, the term “particles” will be used to describe undesirable multimers of protein over a wide range of sizes including submicrometer and micrometer size ranges. Despite being an area of interest to regulatory health agencies, accurate quantification and identification of particles © 2012 American Chemical Society

Received: May 24, 2012 Accepted: June 28, 2012 Published: July 13, 2012 6833

dx.doi.org/10.1021/ac300976g | Anal. Chem. 2012, 84, 6833−6840

Analytical Chemistry

Article

inside a suspended cantilever. Particles that pass through the suspended microchannel are detected due to a momentary shift in the resonant frequency of the suspended cantilever associated with the change in mass caused by the passage of a particle of differing density than the solution. Further, since silicone oil has a density lower than that of water and therapeutic proteins generally have densities greater than water, the direction of the frequency shift clearly distinguishes particles that are positively buoyant (e.g., silicone oil droplets) and negatively buoyant (e.g., protein particles). This method has the potential to be used as an orthogonal approach that can overcome some of the challenges associated with other commonly used subvisible particle characterization methods because the method of detection relies on differences in density rather than differences in optical contrast, conductivity, and/or Brownian motion. The SMR was first described by Burg et al.27 and Godin et 28 al to measure biomolecular adsorption, single cells, and single nanoparticles in aqueous solutions and has since been applied to detect agglutination of microspheres,29 to monitor the growth of single cells30,31 and to determine the density of cells.32 A larger difference between solution density and particle density results in larger buoyant mass and therefore a larger change in resonant frequency. Thus, this method has the potential to determine the size of subvisible particles or, alternatively, the density of a population of particles. As part of method development and assessment, we compare the SMR method with other techniques that are being explored for the routine characterization of subvisible particle content in biopharmaceutical preparations. The comparison is performed using a model system developed for the purpose and consisting of a mixture of differently sized polymer microspheres as well as through the analysis of high protein concentration monoclonal antibody formulations that are filled into and stored within glass prefilled syringes.

undercount or undersize particles in higher concentration protein solutions, which have refractive indices closer to that typical of protein particles.16 In particular, the quantification of smaller particles (0.35 μm >0.8 μm >1.5 μm >2.5 μm >4 μm

5.80 4.99 4.17 2.65 1.27

± ± ± ± ±

0.08 0.06 0.07 0.11 0.08

0.965 1.913 3.121 5.049

± ± ± ±

0.060 0.014 0.012 0.013

Cumulative concentration (106/mL) 6.58 5.59 3.46 1.70

± ± ± ±

0.03 0.03 0.02 0.02

the size ranges in Table 1. Figure 2 plots the mean buoyant mass of each population as a function of solution density.

microspheres, and melamine microspheres; see Supporting Information). Accounting for the differences in accessible size ranges as well as accessible concentration range between SMR, Coulter counter, light obscuration, and flow microscopy, the mean particle sizes for each subpopulation of microspheres agree well, as illustrated in Figure 1. Furthermore, the total particle results obtained by the Archimedes system are in very good agreement with the particle distribution obtained by Coulter counter. While it is noted that the Coulter counter cannot reliably quantify particles as small as 500 nm, the percentage difference in the particle concentrations obtained by the two measurement methods is less than 30% for the larger size subpopulations. There are several notable differences between the results obtained by methods that rely on optical contrast (i.e., light obscuration and flow microscopy) and those obtained by techniques that rely on different mechanisms of detection. In addition to the smallest detectable size being limited to approximately 1.5 μm in size, both the light obscuration and flow microscopy methods show lower counts overall and higher polydispersity in particle size for each subpopulation of particles. The relatively large polydispersity relative to SMR and Coulter counter is attributed to the lower resolution of the optical techniques to distinguish between particles in this size range. Further, it was found that changing detection parameters on the FlowCam instrument, such as lowering the detection threshold, leads to an increase in measured particle size because more particles that are slightly out of focus are counted. These particles possess blurry edges that are above the detection threshold, which make accurate sizing difficult. Thus, care must be taken in interpreting these results. Identification of Microspheres by Determining Particle Density. To determine the density of the different populations of microspheres in the model system, the mean buoyant mass of each population was measured in three different solutions that consisted of the native stock mix diluted 1:1 with each of the following: water, 1050 mM sucrose, and 2 M sucrose. To capture all of the data and ensure consistency for this complex mixture, the mean buoyant mass for each population was determined by averaging the buoyant mass of all of the particles in fixed size ranges for both the positively and negatively buoyant microspheres with populations defined by

Figure 2. Particle densities of the different microsphere populations were determined by extrapolating the buoyant mass of each population to neutral density.

Extrapolation of these data to neutral density results in a density of 1.048 ± 0.001 g/mL for the polystyrene particles, which is within 0.2% of the expected value of 1.050 g/mL. Note that the smallest size polystyrene particle (d = 565 nm) did not produce a large enough signal relative to baseline noise to be reliably detected when the stock mix solution was diluted with water. Additional experiments (data not shown) demonstrate that a density difference as little as 5% is sufficient to reliably detect 565 nm particles (i.e., 565 nm polystyrene microspheres in water). Extrapolation of the mean buoyant mass of the melamine microspheres resulted in a density of 1.455 ± 0.017 g/mL, based on the two different size populations of melamine microspheres in the stock mix. While the value obtained is within 3.7% of the expected value of 1.510 g/mL, the percentage error for melamine microspheres was significantly 6837

dx.doi.org/10.1021/ac300976g | Anal. Chem. 2012, 84, 6833−6840

Analytical Chemistry

Article

limit of approximately 580 nm for silicone oil droplets and approximately 375 nm for protein particles if densities of 0.969 and 1.372 g/mL, respectively, are assumed. It is interesting to observe an apparent dropoff in concentration of particles at sizes closest to the threshold of detection. This observation is ascribed to decreasing probability of detection for particle sizes very close to the detection threshold as the particle signal approaches baseline noise. To evaluate this, a thermally stressed mAb 1 sample was analyzed by SMR with both a standard Micro sensor chip and a prototype Nano sensor chip, which allows the analysis of particles in a smaller size range at the expense of requiring higher particle concentrations. The particle distributions obtained by the two sensors showed relatively good agreement for particle sizes away from the lower limit of detection for the Micro sensor, but the Nano sensor shows that the concentration of particles continues to increase as the size range approaches the theoretical lower limit of the Micro sensor (see Supporting Information). These data sets suggest that a conservative threshold of approximately 4−5 times the baseline noise will avoid the appearance of the apparent drop-off at the expense of not being able to detect the smallest particles. For a sample exhibiting similar baseline noise as that shown in Figure 3, increasing the detection threshold to this multiple would result in an increase in smallest detectable particle size to ∼700−750 nm for silicone oil droplets and ∼460−495 nm for protein particles. It was also qualitatively observed that baseline noise was sample-dependent, with buffer solutions exhibiting slightly lower noise than protein solutions. It is hypothesized that solutions with higher baseline noise possess higher amounts of subvisible particle content slightly below the limit of reliable detection, giving rise to an apparently noisier baseline. Table 3 presents a comparison of the mAb sample analyzed by light obscuration and flow microscopy with the results obtained by SMR. The average volume of mAb 1 sample analyzed per run was 2.6 μL, with an average run time of approximately 3.65 h. Each technique has relatively high precision in size ranges where high particle counts are observed, as demonstrated by standard deviations that are typically within 5−15% of the mean values. While flow microscopy generally has been observed to produce higher counts than light obscuration due to differences in sensitivity,13,14 results obtained by SMR show a particle concentration that is between those reported by the two optical techniques for the size ranges that overlap. The SMR results represent an average of three

higher than that obtained for polystyrene microspheres. The higher error is likely due to the larger extrapolation to neutral density. In addition, the slightly higher uncertainty in mean buoyant mass of melamine particles in the 0.3−2.5 μm range also contributed to the difference between the calculated and expected densities. Distribution of mAb 1. To evaluate the potential application of the SMR method to characterize micrometer and submicrometer particles in representative protein solutions, prefilled syringe samples of monoclonal antibody 1 (mAb 1) formulated at 150 mg/mL were analyzed. A typical distribution of particles from a freshly expelled sample sorted by relative buoyancy is presented in Figure 3. The solution density was

Figure 3. Typical particle distribution for monoclonal antibody 1 in a glass prefilled syringe.

found to be 1.055 g/mL and the majority of the measured particles have a density less than that of the protein solution. The total concentration of particles was approximately (2.01 ± 0.19) × 105 particles/mL, with greater than 97% of detected particles having density lower than that of the solution. Significantly lower particle concentration was observed for samples stored in vials (data not shown). The relative density differences give rise to different lower limits of detection for silicone oil droplets and protein particles. A detection threshold of 0.07 Hz corresponds to a theoretical smallest detectable size

Table 3. Comparison of Cumulative Particle Concentrations for Mab 1 by Light Obscuration, Flow Microscopy, and SMRa Cumulative concentration (particles/mL) Light obscuration size range >350 nm >500 nm >800 nm >1.0 μm >2.0 μm >5.0 μm >10 μm >25 μm

mean

735 133 19 0

std dev

6 10 4 0

Flow microscopy mean

13 100 2010 180 1

std dev

600 130 60 3

Positively buoyant SMR mean 196 000 149 000 75 000 5000