Absolute Quantitation of Intact Recombinant Antibody Product Variants

Subscriber access provided by UNIV OF CAMBRIDGE

Article

Absolute Quantitation of Intact Recombinant Antibody Product Variants Using Mass Spectrometry Frank D. Macchi, Feng Yang, Charlene Li, Chenchen Wang, Anh Nguyen Dang, Joseph C. Marhoul, Hui-Min Zhang, Timothy Tully, Hongbin Liu, X. Christopher Yu, and David Michels Anal. Chem., Just Accepted Manuscript • DOI: 10.1021/acs.analchem.5b02627 • Publication Date (Web): 16 Sep 2015 Downloaded from http://pubs.acs.org on September 17, 2015

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a free service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are accessible to all readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

Analytical Chemistry is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 36

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Analytical Chemistry

Absolute Quantitation of Intact Recombinant Antibody Product Variants Using Mass Spectrometry

Frank D. Macchi†,‡, Feng Yang*,†,‡, Charlene Li†, Chenchen Wang§, Anh Nguyen Dang†, Joseph C. Marhoul†, Hui-min Zhang†, Timothy Tully¶, Hongbin Liu†, X. Christopher Yu†, David A. Michels†

†

Protein Analytical Chemistry, Genentech Inc., 1 DNA Way, South San Francisco, California

94080, United States §

Eurofins Lancaster Laboratories, Inc., 2425 New Holland Pike, Lancaster, Pennsylvania 17601,

United States ¶

Purification Development, Genentech Inc., 1 DNA Way, South San Francisco, California

94080, United States

‡

Co-first authors

*Corresponding author: Email: [email protected]; Fax: +1 650 225 3554; Phone: +1 650 467 8190.

1 ACS Paragon Plus Environment


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ABSTRACT Accurate and precise quantitative measurement of product-related variants of a therapeutic antibody is essential for product development and testing. Bispecific antibodies (bsAbs) are Abs composed of two different half-antibody arms, each of which recognizes a distinct target, and recently they have attracted substantial therapeutic interest. Due to the increased complexity of its structure and its production process, as compared to a conventional monoclonal antibody, additional product-related variants, including covalent and non-covalent homodimers of half antibodies (hAbs), may be present in the bsAb product. Sufficient separation and reliable quantitation of these bsAb homodimers using liquid chromatography (LC) or capillary electrophoresis-based methods is challenging because these homodimer species and the bsAb often have similar physicochemical properties. Formation of non-covalent homodimers and heterodimers can also occur. In addition, since homodimers share common sequences with their corresponding halves and bsAb, it is not suitable to use peptides as surrogates for their quantitation. To tackle these analytical challenges, we developed a mass spectrometry-based quantitation method. Chip-based nanoflow LC time-of-flight mass spectrometry coupled with a standard addition approach provided unbiased absolute quantitation of these drug-product-related variants. Two methods for the addition of known levels of standard (multi- or single-addition) were evaluated. Both methods demonstrated accurate and reproducible quantitation of homodimers at the 0.2% (w/w) level, with the single-addition method having the promise of higher analytical throughput.


Page 2 of 36

Page 3 of 36

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


INTRODUCTION Biologics such as recombinant antibodies (rAbs) represent one of the most rapidly growing product classes of protein therapeutics. One type of next-generation rAb, the bispecific antibody (bsAb), recognizes two distinct targets and has attracted substantial therapeutic interest in recent years.1-4 One efficient way to generate a bsAb is through knob-into-hole technology, whereby complementary mutations are made in the CH3 domain of each heavy chain of the antibody to form knobs and holes.5, 6 Compared to conventional rAbs, bsAbs have additional product-related variants that should be considered, such as covalent and non-covalent homodimers of knob or hole halves and non-covalent heterodimers of the bsAb halves. (Homodimers and heterodimers in this context are the combination of two hAbs, not to be confused with dimers composed of two full antibodies). The potential bsAb variant forms are shown in Figure 1. These productrelated variants, either knob or hole homodimers in bsAbs, have the potential to cause undesired toxicity and/or immunogenicity in the clinic. Therefore, it is important to obtain detailed characterization and accurate quantitation of bsAb homodimer variants to support process development and appropriate product quality testing to ensure product safety. Analytical tools such as liquid chromatography (LC)7-10 and capillary electrophoresis (CE) have been successfully used to separate and quantify size and charge variants of conventional rAbs. However, it can be a challenging task to obtain sufficient separation and reliable quantitation of homodimers with these methods. First, it is difficult to measure bsAb variants when the variants have highly similar physicochemical properties compared to the unmodified rAb, thus LC and CE methods are often found to be inadequate for separation and accurate quantitation of these species. For example, it is difficult to differentiate homodimers11 from the bsAb due to poor resolution using size-exclusion chromatography (SEC). Another level of



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

complication comes from the fact that there are pH-dependent distributions of conformational isomers/structures of halves and homodimers, formation of non-covalent homodimer11 and heterodimer of halves, and all types of other non-covalent interactions among halves, homodimers, and bsAbs. Many of these challenges are observed in different bsAb products that are currently under our investigation. Therefore, there is a great need to develop alternate approaches to reliably quantify bsAb homodimers. Fortunately, high sensitivity and highresolution mass spectrometry (MS)-based methods performed under denaturing conditions can dissociate non-covalent interactions and differentiate covalent and non-covalent homodimers of either half from bsAb by mass, providing an opportunity to tackle these challenges. The common internal standard-based MS protein quantitation approaches typically use proteolytic peptides as surrogates for the protein of interest.12-17 However, peptide level quantitation is not suitable for homodimer quantitation because homodimers share common sequences with their corresponding halves and bsAbs. In addition, it is hard to demonstrate quantitative accuracy using this method due to imperfect proteolytic cleavage efficiency and peptide loss from sample preparation.18, 19 When considering variations of sample processing across replicates and technical variations of data acquisition, a CV of ≤ 20% is usually considered as acceptable with the use of an external calibration curve.18 For intact protein quantitation, the external calibration approach can also be applied. However, when the sample matrix (components of a sample other than the analyte of interest) varies from analysis to analysis, the MS response and the slope of the calibration curve can change and create a quantitation accuracy bias. To mitigate these challenges, a standard addition approach can be used, which is usually performed by separately adding several different amounts of analyte to individual aliquots of the


Page 4 of 36

Page 5 of 36

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


test sample; measurement is followed by extrapolation of the standard addition curve to zero response. 20-25 However, the low analysis throughput makes this method less appealing to use. In addition, because the standard addition method involves extrapolation from known data, it is often regarded as less precise than interpolation-based measurements.20 These two limitations of the standard addition approach can be mitigated by using only a single addition of the largest possible amount of standard that falls within the linear range of response.26, 27 This singleaddition approach increases analytical throughput as compared to a multi-point standard addition method. It was previously reported that better precision can be achieved in estimating the calibration slope of a standard addition linear regression curve when the measurements are confined to the ends of the linear concentration range.26, 27 To our knowledge, application and evaluation of multiple and single standard addition methods has not been reported for MS-based protein quantitation of intact antibodies. In this work, we describe the novel application of the standard addition approach using Chipbased nano-flow reversed-phase liquid chromatography coupled with in-line time-of-flight mass spectrometry (nRPLC-Chip-TOFMS)28, 29 to obtain unbiased absolute quantitation of homodimers in bsAbs. In the case of bsAb1, a Chinese hamster ovary (CHO) cell-derived IgG1 bsAb, knob-knob homodimers can potentially trigger non-specific signaling pathways. Due to this clinical safety concern, knob homodimers may need to be quantified with adequate sensitivity (for example, ≤ 1% for bsAb1). Because of the knob design, combined with effective assembly and purification processes, there was no MS-detectable covalent knob homodimers in the bsAb1 product. Therefore, we focused on quantitation of non-covalent knob homodimers in bsAb1 under denaturing conditions of LC-MS analysis, using knob hAb as the surrogate. By doing so, the worst-case scenario for clinical safety concerns was considered, in which all knob



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

hAbs formed non-covalent homodimers. The accuracy and precision of different methods of standard addition (multi- and single-addition approaches) were evaluated and compared using the bsAb1.

EXPERIMENTAL SECTION Materials and Reagents. Knob and hole hAbs for bsAb1 (a CHO cell-derived, nonglycosylated IgG1 bsAb) were each produced separately in CHO cells, purified by Protein-A after individual expression, mixed, and assembled into bsAb 1 using reduced glutathione disulfide exchange at Genentech Inc. (South San Francisco, CA).30 Assembled bsAb1 was further purified using several LC steps. Unassembled knob-hAb standard (KnobStd, approximately 98% purity as assessed by both SEC and RPLC) was also produced and purified at Genentech Inc. The KnobStd is the same species that is to be quantified in the test sample. Covalent knob homodimers was estimated to be 0.2% or less in the KnobStd by MS analysis. LC-MS grade mobile phase A (0.1% formic acid in water) and mobile phase B (0.1% formic acid in acetonitrile) were both purchased from Honeywell Burdick & Jackson (Muskegon, MI).

Sample Preparation. Both test samples and the KnobStd were diluted to 1 mg/mL using LC-MS grade water. To assess assay linearity, ten levels (from 0.1 to 99 µg/mL) of KnobStd were spiked into the same amount of bsAb1. First, the 0 and 99 µg/mL-spiked samples were generated, with each containing the same amount and concentration of bsAb1. The other spike level samples were made by blending samples with 0 and 99 µg/mL of KnobStd spike levels to minimize pipetting errors. This blending method was also used to prepare five different levels (0, 1, 4, 9, and 44 µg /mL) of KnobStd-spiked test samples to perform the multi-addition assay,


Page 6 of 36

Page 7 of 36

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


through co-mixing test samples with the lowest (0 µg /mL) and highest (44 µg/mL) spike levels. For the single-addition method, only the 0 and 50 µg/mL of KnobStd-spiked test samples were prepared. Each of the KnobStd-spiked bsAb1 samples were then diluted 10-fold (v/v) to approximately 0.1 mg/mL with 5% acetonitrile in 0.1% formic acid solution prior to nRPLCChip-TOFMS analysis.

nRPLC-Chip-TOFMS Analysis. Each sample was analyzed by liquid chromatography electrospray ionization mass spectrometry (LC-ESI-MS) using an Agilent 6210 or 6230 time-offlight mass spectrometer coupled with a nano-Chip LC-ESI source. About 5 ng of protein sample (0.05 µL injection volume) was injected onto a custom Chip with a 40 nL trap column for desalting and an equivalent analytical column, 43 mm × 75 µm, Zorbax 300SB-C8 (5 µm, Agilent Technologies, Santa Clara, CA) for separation at 400 nL/min. The mobile phases A and B used were 0.1% formic acid in water and 0.1% formic acid in acetonitrile respectively.29 A gradient from 25% to 90% solvent B in 6 min was applied to all samples. The mass range was set to 200–3200 m/z. For the multi-addition method, injection of a set of five different spike level bsAb1 samples was repeated to minimize the effect of run-to-run LC-MS analysis variation on quantitation. For the single-addition method, each sample set consisted of triplicate injections of each of 0 and 50 µg/mL KnobStd-spiked bsAb1 samples.

LC-MS Data Extraction, Deconvolution, and Normalization. Agilent MassHunter Quantitative Analysis Workstation Software version B.04.00 for Chip-TOF MS model 6210 and version B.06.00 for MS model 6230 were used for MS data extraction and deconvolution. MS spectra of all data in a standard addition experiment were extracted in the selected 5-minute total



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ion chromatogram (TIC) window to ensure the entire peak covering the knob hAb elution region (especially the beginning portion of the TIC which contains the majority of knob hAb) was included (Figure 2A). The molecular masses were derived from deconvolution of MS spectra using the MassHunter maximum entropy algorithms. The deconvolution parameters used for the knob hAb included a mass range of 60,000–80,000 Da, limited m/z range of 1400–2100, signal to noise ratio of 20, average mass % peak height of 25%, minimum consecutive charge states of 5, and minimum protein fit score of 8. The mass and m/z range for deconvolution was selected to provide a balance of specificity, sensitivity, and minimized deconvolution artifacts on quantitation. To minimize the effect of run-to-run LC-MS analysis variations on quantitation, the deconvoluted intensity of knob hAb peak was normalized prior to standard curve plotting. First, the intensity of the most intense charge envelope peak of bsAb1 (Peak1) from each MS spectrum (Figure 2B) and intensity of the deconvoluted knob hAb peak, 71745 Da (±50 ppm) (Figure 2C) were recorded. Since the same amount of bsAb1 was present in all spike samples in the standard addition method experiment design, it was feasible to use the intensity of bsAb1 Peak1in the MS spectrum to normalize the deconvoluted knob hAb data. After normalization, the average of the deconvoluted knob hAb intensities from injection replicates was taken as the analytical response for each spike level sample.

Standard Addition Curve Generation and Knob hAb Quantitation. To calculate the level of knob hAb in the testing sample, the standard addition curve was generated by plotting the analytical response for each spike level sample vs. KnobStd (µg/mL) spike levels, and the linear regression equation y=mx+b was determined. The concentration of knob hAb in zero-spiked


Page 8 of 36

Page 9 of 36

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


sample (µg/mL) is equal to b/m [y intercept (b) divided by slope (m)]. Therefore, the concentration of knob hAb in the diluted testing sample (1 mg/mL) is b/m multiplied by the sample dilution factor in the standard addition step. The % (w/w) of knob hAb is equivalent to the calculated concentration of knob hAb (µg/mL) in the diluted 1000 µg/mL bsAb1 test samples divided by 10. This % (w/w) value was reported for bsAb1 since drug administration dosage is based on the amount of bsAb1 drug per kg of patient weight.

RESULTS AND DISCUSSION Measurement of Knob hAb As the Surrogate for Non-covalent Knob Homodimer in bsAb1. Covalent or non-covalent knob homodimers in bsAb1 are undesired product variants. Efforts had been made to separate and quantify these homodimers using various LC and CEbased methods, including RPLC, hydrophobic interaction chromatography, cation exchange and anion exchange chromatography, SEC, imaged capillary isoelectric focusing, and capillary electrophoresis-sodium dodecyl sulfate. However, none of these methods were successful in separating knob homodimers from bsAb1 with good resolution, sensitivity, and recovery (data not shown), due to several factors. First, homodimers and bsAb1 have similar physicochemical properties such as size and charge. Second, non-covalent homodimers and heterodimers have been observed under different experimental conditions (Figure 1). Most of these analytical challenges were also observed in the different bsAb products that are currently under investigation. Since there were no MS-detectable covalent knob homodimers in the final bsAb1 product, we focused on quantitative measurement of the non-covalent knob homodimers. Under the denaturing conditions of nRPLC-Chip-TOFMS analysis, non-covalent knob homodimers were



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

dissociated into knob hAb monomers, as were non-covalent knob-hole heterodimers and other non-covalent complexes formed with the knob hAb in the bsAb1 product. In this work, we used the total amount of measured knob hAbs as a surrogate for the non-covalent knob homodimer, assuming all measured knob hAbs originated from non-covalent knob homodimers. We chose this as a worst-case scenario for clinical safety concerns. Our preliminary results indicated that the knob hAb co-eluted with the hole hAb, and could not be completely resolved from bsAb1 by Chip-based RPLC, or even by reversed-phase ultraperformance liquid chromatography (data not shown). Therefore, for a robust, fast, and sensitive analysis, we chose to use the highly stable Chip-based nano-flow LC coupled to TOF-MS28 in this study. An LC gradient was designed so that all the species in the bsAb1 samples were eluted as one broad peak (at about 5 min, Figure 2A). The representative raw and deconvoluted MS spectra of knob hAbs are shown in Figure 2B and 2C, respectively. We first attempted to apply the working curve external calibration approach coupled to nRPLC-Chip-TOFMS for the quantitation, where a single calibration curve was used to analyze multiple samples. However, when using the calibration curve method, we found that the MS responses (slope of the working curve) of the knob hAb varied as much as 24% between samples with only minor matrix differences. To address these analytical challenges, in this study we developed the standard addition method coupled to nRPLC-Chip-TOFMS for the unbiased quantitative measurement of homodimers. The general workflow of the standard addition method is shown in Figure 3. First, a set of samples with equal volumes was prepared, which contained exactly the same amount/concentration of bsAb1 testing sample but different levels of spiked KnobStd. Therefore, the matrices for samples with different spike levels were exactly the same, minimizing the matrix


Page 10 of 36

Page 11 of 36

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


effect on quantitation accuracy. Examples of sample preparation for a standard addition experiment are also shown in Figure 3. The spiked bsAb1 samples were analyzed by nRPLCChip-TOFMS. MS spectra across the entire protein elution peak were deconvoluted to ensure complete recovery of knob hAb. The peak intensity of deconvoluted mass spectra of knob hAb was then used to generate the standard addition curve. The level of knob hAb (µg/mL) in the 1 mg/mL bsAb1 test sample was calculated from the linear regression equation determined for the standard addition curve, as shown in Figure 3. The level of knob hAb in bsAb1 was reported as % (w/w) for the testing sample, as detailed in the Experimental Section. In this study, while the absolute quantitation of the knob hAb, as the surrogate for non-covalent knob homodimer of bsAb1, was used as a proof-of-concept application for the standard addition approach, we expect that the same method can be applied to covalent or non-covalent knob and hole homodimers, or other product-related variants in other bsAbs or conventional rAbs, as long as these variants have distinct masses and good resolution from the product and other minor variants. Briefly, depending on whether the minor product variant has lower or higher mass than the antibody product, mass differences of several hundred Da may be needed to resolve the minor product variant from antibody product and other product variants. Baseline resolution will be ideal, while additional sample preparation steps, such as antibody heavy chain C-terminal lysine removal, can be taken to reduce the heterogeneity of the product to improve the resolutions.

Linearity and Measured Level of Knob hAb in bsAb1. To test the linearity of the assay and to establish the level of knob hAb in bsAb1 sample for further testing, 11 bsAb1 samples were prepared, which included 10 that were spiked with various (0.1–99 µg/mL) KnobStd levels and the zero-spiked samples. The overlaid deconvoluted mass spectra of knob hAb for the set of



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

11 samples are shown in Figure 4A. The MS responses represented by the deconvoluted intensity of knob hAb at the peak apex (71745 Da ±50 ppm) increased with the spike level, and were linear within the tested range with excellent correlation (R2 = 0.9997) (Figure 4B). In order to assess the analytical response “recovery” at each addition level, the theoretical response was calculated using the linear regression equation and then compared to the measured responses. All recovery data were within 92–105% of the theoretical value, demonstrating that the analytical response (deconvoluted intensity) is accurate throughout the linear concentration range tested (data not shown). The calculated amount of knob hAb present in the bsAb1 material was 0.2% (w/w) from the linearity experiment (Figure 4). Re-analyzing the data with limited, multi-addition range (the lower end, from 0 to 49 µg/mL spike) resulted in same knob hAb value in bsAb1, further supporting the assay linearity within the tested range. For subsequent standard addition method experiments performed in this study, the upper limit of knob hAb spike level was set to be 50 µg/mL [corresponding to 5% (w/w) of knob hAb in bsAb1 sample] to make sure that future bsAb1 samples that might contain a high level [up to 5% (w/w)] of knob hAb, when spiked, are still within the established assay linearity range.

Evaluation of Assay Accuracy In Samples With a Known Level of Knob hAb. To further demonstrate assay accuracy of the standard addition methods, bsAb1, containing 0.2% (w/w) level of knob hAb, was spiked with different levels of KnobStd to create testing samples with various [0.3–5.2% (w/w)] levels of knob hAb. The recovery for all tested samples using both multi- and single-addition methods was excellent, and the variation was within 10% of the expected level, indicating high assay accuracy. The correlation between the expected and


Page 12 of 36

Page 13 of 36

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


measured knob hAb levels in all tested samples is shown in Figure 5. Linearity was achieved for both correlation curves, with R2>0.99. The slopes of both correlation curves were close to 1. These results demonstrate that both standard addition methods can be used for accurate absolute quantitation of knob hAb in bsAb1.

Limit of Detection and Limit of Quantitation. Next, we used two approaches to estimate the limit of detection (LOD) and limit of quantitation (LOQ) values for the standard addition method for knob hAb, a surrogate for non-covalent knob homodimers of bsAb1. First, the LOD and LOQ were calculated using the equation from the ICH Q2B guideline: LOD=3.3(σ/S) and LOQ=10(σ/S),31 respectively, where σ was the standard deviation of knob hAb response in zerospiked bsAb1 reference material (n=6, process replicates), and S was the averaged slope of the standard addition curve. The calculated LOD and LOQ for knob hAb measurements were 0.06% and 0.19% (w/w), respectively. Another reported strategy was also applied to estimate the LOQ of the assay, based on a signal-to-noise ratio (S/N) ≥10 and measurement coefficients of variation (CV)

Absolute Quantitation of Intact Recombinant Antibody Product Variants

Recommend Documents