Orthogonal pre-use and post-use efficiency testing for single-use anion exchange chromatography

Three efficiency tests for single-use AEX chromatography devices have been developed and applied to six capsule formats of a new, salt tolerant, single-use AEX product. All the tests have been designed to be performed with simple equipment and common reagents.

Journal of Chromatography A 1654 (2021) 462445 Contents lists available at ScienceDirect Journal of Chromatography A journal homepage: www.elsevier.com/locate/chroma Orthogonal pre-use and post-use efficiency testing for single-use anion exchange chromatography Jonathan F Hester a,∗, Xinran Lu a, Jacob D Calhoun a, Rebecca A Hochstein a, Eric J Olson b a b 3M Separation and Purification Sciences Division, 3M Center 236-1C-14, St Paul, MN 55144-1000, United States 3M Corporate Research Analytical Laboratory, 3M Center 201-BS-03, St Paul, MN 55144-1000, United States a r t i c l e i n f o Article history: Received April 2021 Revised 12 July 2021 Accepted 22 July 2021 Available online 30 July 2021 Keywords: Single-use chromatography Anion exchange (AEX) chromatography Polishing chromatography Monoclonal antibodies (MAbs) Downstream processing Viral clearance a b s t r a c t Three efficiency tests for single-use AEX chromatography devices have been developed and applied to six capsule formats of a new, salt tolerant, single-use AEX product All the tests have been designed to be performed with simple equipment and common reagents By performing each of the three tests on undamaged capsules and capsules intentionally damaged with small defects, in tandem with Phi-X174 challenges in a high-salt buffer, relationships between test results and viral clearance have been obtained A pre-use pressure-based installation verification test is simply performed during equilibration of the device and effective at identifying gross bypass defects, for example, due to internal seal breakage Passing outcomes of a post-use installation validation bubble point test are associated with ≥ log reduction value (LRV) of viral clearance A new, non-destructive, pre-use AEX capacity test involves challenging the device with chloride ions and is orthogonal to the other two tests in that it can detect chemical defects, as well as mechanical ones Passing outcomes of this test correspond to > LRV viral clearance and provide in situ assurance of the expected AEX dynamic capacity prior to use Selection of a pair of preuse and post-use tests can provide robust risk reduction with respect to viral clearance by single-use AEX devices in biopharmaceutical purifications © 2021 3M Company Published by Elsevier B.V This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/) Introduction Flow-through anion exchange (AEX) chromatography is frequently used in biopharmaceutical purification processes for reduction of net-negatively charged host cell proteins (HCPs) and viral reduction as part of a validated viral clearance strategy [1,2] AEX column chromatography is the technology most often used for electrostatic viral clearance, particularly in commercial scale biopharmaceutical manufacturing, where columns have established a long history of reliable and well understood performance [3] Still, validation of HCP and viral clearance by AEX columns in biopharmaceutical processes involves complexities which contribute significantly to operational and regulatory costs Manufacturers must be concerned with the possibility of micro-channeling in columns which may result from defects in column packing, a concern routinely mitigated by in situ measurement of the asymmetry of the elution peak resulting from the upstream pulse injection of an analyte and quantification of the height equivalent to a theoretical ∗ Corresponding author E-mail address: jfhester@mmm.com (J.F Hester) plate (HETP) derived from the elution peak retention time and breadth [4–6] Another concern is the potential for loss of viral clearance with resin re-use, which may extend over hundreds of use cycles with intervening cleaning procedures that have the potential to cause resin degradation [6–8] An assessment of the effect of resin re-use on viral clearance is thus generally recommended on a product-by-product basis [1,6,7] In recent years, the introduction of single-use AEX technologies has illuminated the potential for reduced regulatory and operational costs associated with flow-through AEX chromatography [5,9–11] Physically resembling and operated like filters, singleuse AEX products benefit from improved specific capacity and enhanced flow rates compared with columns due to the replacement of diffusive kinetics with convective flow These features have led researchers to note the potential for simpler operation, decreased processing times, and reduced buffer consumption leading to improved economics relative to columns Additionally, their singleuse nature obviates validation costs associated with cleaning and performance over repeat use cycles, including viral clearance performance While single-use AEX products were initially used primarily in laboratory and process development activities, recent ad- https://doi.org/10.1016/j.chroma.2021.462445 0021-9673/© 2021 3M Company Published by Elsevier B.V This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/) J.F Hester, X Lu, J.D Calhoun et al Journal of Chromatography A 1654 (2021) 462445 vances in ligand and media design have resulted in devices with high specific capacity and robust performance across fluid conditions, making them viable alternatives to packed columns in commercial manufacturing [12] With respect to the deployment of single-use AEX devices in large-scale manufacturing, investigators have noted the need for sensitive, in situ test methods capable of detecting device defects that might result in reduced viral clearance HETP and peak asymmetry studies used successfully on columns are not sensitive for single-use devices, given their high flow rates, large mixing volumes, and short bed heights [13] In defining test strategies for single-use devices, it is useful to consider the viral clearance risks that might occur Viral clearance loss could occur due to mechanical bypass resulting from media or seal damage which might originate in manufacturing or during shipping and handling, for example Alternatively, premature viral breakthrough could occur in a mechanically integral capsule due to any of a few possible chemical defects: regions of the media not properly chemically functionalized during manufacturing or missing AEX media layers, for example Further, there is a time distribution of risk during bioprocessing To reduce the risk of processing valuable productcontaining fluid with a damaged capsule, resulting in a process deviation and the need for costly re-processing, a non-destructive pre-use efficiency test is desired However, there is also a risk that capsule damage could occur during processing, for example, due to over-pressurization To mitigate this risk, a post-use efficiency test might be preferred A number of efforts have been made to develop and employ in situ efficiency tests for single-use chromatography devices Diffusion tests are commonly used, wherein the capsule media is prewet with an aqueous solution, a constant upstream gas pressure is applied, and the gas diffusion rate across the media is measured [14,15] These are useful for detecting gross mechanical leaks such as seal damage or holes in the media They would fail to detect chemical defects, however Multiple in situ tests have been described that can characterize both mechanical and chemical integrity of single-use devices Bind-and-elute type tests involve loading of the AEX device with an excess of a binding analyte, washing the media, and then eluting the analyte and characterizing the resulting elution profile [16,17] Previously described breakthrough-type tests involve challenging the device with a specific analyte while monitoring the concentration of the analyte in the filtrate Investigators have developed breakthrough-type tests based on specialized analytes [18] as well as controlled pH changes in the feed solution [19–23] In situ efficiency tests developed to date generally have one or more of the following challenges with respect to application at commercial scale Many of the tests, particularly those of the bindand-elute type, rely on subjective comparison of the analyte elution profile with that of a reference device Quantitative determination of what deviation from the reference profile is indicative of a defect expected to result in viral clearance loss, vs inconsequential media variability, is difficult, and such subjectivity is not generally suitable to a Current Good Manufacturing Practices (cGMP) environment Some of the tests utilize specialized reagents, not generally available in biopharmaceutical manufacturing plants, which would need to be inventoried specifically for the tests Finally, the breakthrough-type tests rely on knowledge of the upstream fluid volume between the fluid injection point and the downstream breakthrough detector This includes the fluid volume within the AEX device as well as any upstream and downstream tubing While such holdup volume estimates are relatively straightforward and commonly used on lab-scale equipment, they are problematic at manufacturing scales, where long lengths of large-diameter tubing may be utilized and where the tubing and capsule headspaces may contain variable volumes of air bubbles, for example Herein are described three in situ efficiency tests for single-use AEX devices from which may be selected an orthogonal set of preand post-use tests suitable for commercial manufacturing Performance of the efficiency tests is assessed in the context of a recently commercialized, hybrid AEX device that combines a quaternary ammonium (Q) functional nonwoven AEX media with a novel guanidinium (Gu) functional AEX membrane, achieving robust HCP and virus reduction over wide ranges of fluid pH and ionic strength [12] The test methods have objective and quantitative pass/fail criteria and utilize reagents commonly available in biopharmaceutical manufacturing environments A non-destructive, pre-use pressure-based efficiency test is conveniently performed during equilibration of the device and is effective in identifying gross mechanical defects (e.g., internal seal damage) that might result in media bypass A new breakthrough-type, pre-use AEX dynamic capacity test [24] can detect fine mechanical or chemical defects, as well as provide in situ measurement of AEX capacity, without the need to estimate holdup volume of the system Finally, a post-use bubble point test identifies any mechanical defect larger than the largest pore size of the AEX membrane pore size distribution We present performance data for each of the three efficiency tests on undamaged single-use devices of a variety of sizes, as well as devices with purposefully introduced defects To facilitate quantitative risk assessment, we present data relating efficiency test outcomes with measured Phi-X174 viral clearance Experimental 2.1 Materials 2.1.1 Reagents Sodium acetate (ACS, anhydrous), potassium chloride (ACS), and Tris base (Biotechnology Grade) were purchased from VWR Sodium chloride (U.S.P.) was purchased from J T Baker N hydrochloric acid solution for buffer pH adjustment was purchased from J T Baker Filling solution for the chloride ion selective electrode (ISE) was purchased from ThermoFisher (Orion ionplus Optimum Results B, item no 90 062) Filling solution for the potassium ISE was purchased from ThermoFisher (Orion ionplus Optimum Results E, item no 90 065) All reagents were used as received All salt solutions were prepared using deionized water provided by a Millipore Milli-Q water purification system which had a resistivity > 18 M •cm 2.1.2 Test capsules 3MTM Polisher ST single-use AEX capsules (3M Company, St Paul, MN) are available in six sizes (denoted BC1, BC4, BC25, BC170, BC340, and BC1020, where the numeral in each designation refers to the nominal media frontal surface area), as detailed elsewhere [12] The capsules contain a hybrid AEX media bed consisting of layers of Q-functional nonwoven, having a combined bed depth of 0.35 cm, and layers of Gu-functional membrane having a combined bed depth of 0.14 cm Efficiency tests were performed on all six capsule sizes Some pre-use AEX dynamic capacity tests were performed using a 2-piece threaded polycarbonate test housing that could restrain a variable number of 25 mm diameter media discs The housing comprised a top piece with a fluid inlet and a vent for removing air from the housing and a bottom piece with a fluid outlet To simulate the media arrangement of a 3MTM Polisher ST single-use capsule, layers of Gu-functional membrane (FM, 3M Company, St Paul, MN) were placed in the downstream portion of the housing A plastic annular ring with an inner diameter of 19 mm and an outer diameter of 25 mm was placed on top of J.F Hester, X Lu, J.D Calhoun et al Journal of Chromatography A 1654 (2021) 462445 pre-use AEX dynamic capacity test on all 3MTM Polisher ST capsule formats A peristaltic pump (Cole-Parmer item no ZM-0752220 with Masterflex L/S Easy-Load II Pump Head, item no 7720060) supplied any of M sodium acetate, 20 mM sodium acetate, or 20 mM KCl solution to the capsule inlet, with valves positioned upstream of the pump as shown to select the challenge solution For BC1, BC4, and BC25 capsules, potassium and chloride ISE’s were mounted in a dual-probe plexiglass flow cell (Item No 791101, FIAlab Instruments, Seattle, WA) with the potassium ISE upstream of the chloride ISE For larger BC170, BC340, and BC1020 capsules, the ISE’s were mounted in a purpose-built polycarbonate flow cell with top and bottom pieces that were joined together and sealed with an o-ring, defining a small internal volume in contact with the ISE tips Upstream and downstream pressure transducers were monitored using a pressure monitor/transmitter (PMAT2A, PendoTECH, Princeton, NJ) The ISE outputs were monitored using a pH/conductivity monitor/transmitter (PDKT-ACCESS-PHCN, PendoTECH) Filtrate mass was monitored using a top-loading scale Mass, ISE output, and pressure data were automatically logged using PendoTECH PressureMAT data logging software (PMATP-GUI, PendoTECH) The capsule was initially filled with M sodium acetate solution, with the vent open, until all air was removed from the capsule, after which the vent was closed With reference to Fig 1, in a first step, bypass valve was directed to bypass and the capsule was flushed with M sodium acetate to standardize the media by replacing all bound counter-ions with acetate In a second step, the capsule was washed with 20 mM sodium acetate to remove from the capsule housing the high acetate concentration from the previous step In a third step, bypass valve was set to direct the filtrate flow through the ISE flow cell and the flow rate of the 20 mM sodium acetate feed solution was reduced to equilibrate the system at the flow conditions of the subsequent challenge step In a fourth step, the capsule was challenged with a 20 mM potassium chloride solution while the downstream ISE responses were monitored Test conditions for each capsule format are detailed in Table Potassium ions are unbound by the AEX media and pass through the capsule with the challenge fluid front Chloride ions are bound by the media, displacing acetate ions Fig is a plot of exemplary resulting sets of potassium and chloride ISE responses for three different 3MTM Polisher ST BC340 capsules The opposing directionality of the K+ and Cl− breakthrough responses is explained by the Nernst equation, wherein the response of an electrochemical cell to an ionic analyte is proportional to the logarithm of the analyte concentration multiplied by a term including the charge of the analyte Assuming collection of the filtrate during the challenge step begins at the moment the feed solution is switched from 20 mM sodium acetate solution to the KCl challenge solution, the potassium ion breakthrough volume is the system holdup volume between the feed solution injection point and the ISE’s The chloride ion breakthrough volume is the volumetric throughput at which the AEX capacity of the device is exhausted The AEX capacity of the device is characterized by the net breakthrough volume (Vnet ), the difference between the chloride and potassium breakthrough volumes This value was normalized by the capsule surface area and expressed in terms of volume of filtrate per unit area (1909 mL / 340 cm2 = 5.6 mL/cm2 for capsule in Fig 2) or microequivalents of chloride per unit area based on the 20 mM challenge solution concentration (112 μequiv/cm2 for capsule in Fig 2) A “pass” criterion of ≥ mL/cm2 (≥ 80 mequiv/cm2 ) was selected on the basis of numerous trials conducted on capsules containing Q-functional nonwoven and Gu-functional membrane media spanning the media specification ranges for AEX capacity Thus, undamaged capsules containing media of both types at the lower specification limit for AEX capacity are expected to pass the above criteria those Next, layers of Q-functional nonwoven (FNW, 3M Company, St Paul, MN) were placed on top of the ring A second plastic annular ring was placed on top of the layers of FNW Finally, an o-ring was placed on top of the media stack and the top portion of the housing was screwed down to restrain the media stack The resulting capsule had an effective frontal media surface area of 2.84 cm2 2.2 Phi-X174 preparation, filtration challenges, and enumeration To prepare viral challenge solutions, a concentrated, filtersterilized Phi-X174 virus stock containing × 1010 plaqueforming units (PFU)/mL was spiked to a target concentration of > × 108 PFU/mL into 50 mM Tris-HCl buffer, pH 8.0, adusted to a conductivity of 20 mS/cm with NaCl After flushing capsules with 50 L/m2 of Tris-HCl, pH 8.0, 20 mS/cm buffer at an areanormalized flow rate of mL/(cm2 -min), the capsules were challenged with 100 L/m2 of the spiked virus solution at the same flow rate The input virus challenge solution and the filtered outputs were enumerated using a plaque assay Samples were serially diluted to a concentration at which countable virus plaques (zones of clearing in a bacterial lawn caused by viral lysis) could be visualized on a 100 mm agar plate (approximately 20–200 plaques/100 μl) 100 μL of the serially diluted virus sample and 50 μl of E coli 13706 (ATCC) overnight host culture were mixed with 2.5 mL of molten top agar (nutrient broth with 0.5% NaCl and 0.9% agar) and poured on top of a 100 mm nutrient agar plate Plates were incubated at 37 °C for 3-4 h and plaques were counted The LRV was calculated by taking the log of the input virus concentration minus the log of the output concentration Viral clearance results measured by this assay were expected to be accurate within approximately ±0.5 LRV [25] 2.3 Pre-use pressure-based installation verification test For BC170, BC340, and BC1020 capsules, a non-destructive, preuse, pressure-based installation verification test was performed as specified by the manufacturer [26] Briefly, the pressure drop across the capsule was monitored as it was flushed with 20 mM sodium acetate at an area-normalized flow rate of 600 L/(m2 -h) A pressure drop greater than or equal to psid was considered a “pass,” while a pressure drop less than psid was considered a “fail” possibly indicating a bypass within the capsule To facilitate higher-rate testing for BC1, BC4, and BC25 capsules, the pressure drop was monitored as the capsule was flushed with 20 mM sodium acetate at an area-normalized flow rate of 1,800 L/(m2 -h) The measured pressure drop was then divided by and the resulting value was compared with the ≥ psid test criterion (Pressure drop was found to be linear with flux within this range of flow rates.) 2.4 Pre-use AEX dynamic capacity test 2.4.1 Pre-use AEX dynamic capacity tests on capsules of all sizes using a PendoTECH system A non-destructive, pre-use efficiency test was conducted on capsules by standardizing the AEX media with bound acetate counter-ions and then challenging the capsules with dilute potassium chloride [24] While, in the case of an undamaged singleuse capsule, this test fundamentally measures the charge density of the AEX media within, it is termed a “dynamic capacity” test because it is advantageously performed under flow conditions identical to those used when the capsule performs viral clearance in use as recommended by the manufacturer Fig is a schematic of a tabletop setup that was used to perform this J.F Hester, X Lu, J.D Calhoun et al Journal of Chromatography A 1654 (2021) 462445 Fig Schematic diagram of setup used to perform pre-use AEX capacity tests on all capsule formats Solid lines denote fluid (tubing) connections, while dashed lines denote electrical connections Fig Exemplary pre-use AEX capacity test data for three 3MTM Polisher ST BC340 (340 cm2 ) capsules [24] Open symbols are potassium ISE responses (left axis) while closed symbols are chloride ISE responses (right axis) J.F Hester, X Lu, J.D Calhoun et al Journal of Chromatography A 1654 (2021) 462445 Table Steps and settings for performing the pre-use integrity test on any capsule size using a PendoTECH control and data acquisition system Numerals in black circles reference the corresponding labels in Fig Step No Setting Capsule Type BC1 Setup L/S® tubing size 14 ISE flow cell BC4 BC25 BC170 BC340 BC1020 14 14 25 25 25 Small scale Capsule filling Flow rate, mL/min Load Feed fluid M sodium acetate Flow path Valve ❻ to bypass Wash Pilot scale 20 500 12 80 500 1000 1000 Volume, mL 12 80 300 500 1000 Feed fluid 20 mM sodium acetate Valve ❻ to bypass Flow rate, mL/min 10 60 Volume, mL 15 45 60 Feed fluid Flow rate, mL/min Volume, mL 1000 1700 1700 4000 Valve ❻ to ISE flow cell 25 170 340 1000 ≥3 ≥ 10 ≥ 50 ≥ 500 ≥ 500 ≥ 1000 20 mM potassium chloride Valve ❻ to ISE flow cell Flow path End criteria 1000 Feed fluid Flow rate, mL/min 500 20 mM sodium acetate Flow path Challenge 500 Flow rate, mL/min Flow path Equilibrate 500 25 170 340 1000 End when significant Cl− ISE breakthrough response is observed 2.4.2 Pre-use AEX capacity tests on small capsules utilizing a Cytiva ÄKTA chromatography system Pre-use AEX dynamic capacity tests on media configurations in the polycarbonate test housing were performed using a Cytiva ÄKTA avant 25 chromatography system (Item No 28930842, Cytiva) Potassium and chloride ISE’s were mounted in a dual-probe plexiglass flow cell (Item No 791101, FIAlab Intsruments, Seattle, WA) with the potassium ISE upstream of the chloride ISE The resulting ISE probe assembly was mounted on the ÄKTA chromatography system with HPLC tubing from the “Out1” position of the ÄKTA outlet valve directed to the inlet of the side of the flow cell containing the potassium ISE and tubing from the opposite side of the flow cell directed to a waste container The ÄKTA system was outfitted with a Cytiva I/O-box E9 (Item No 29011361, Cytiva) to enable automatic data logging of the ISE outputs Input to the I/O-box E9 was provided by two analog cables with female jacks (Item No 290-1009-ND, Digi-Key) The ISE’s were connected to the input cables using two BNC female interconnect jacks (Item No ARF1069-ND, Digi-Key) Automated data collection from the ÄKTA system was provided by Cytiva UNICORN software, which was configured to log I/O-box E9 inputs according to the manufacturer’s instructions Deionized water was setup on system pump “A” at position “A1.” A 0.5 M sodium acetate equilibration solution was setup on system pump “B” at position “B1.” The 20 mM KCl challenge solution was setup on the sample pump at position “S1.” Generally, Cytiva ÄKTA avant and ÄKTA pure chromatography system models are suitable for performing the pre-use AEX dynamic capacity test on 3MTM Polisher ST BC1, BC4, and BC25 capsules The polycarbonate housing was assembled as described in Section 2.1.2 and the resulting filter assembly was mounted on a column position of the ÄKTA system With the vent open, 0.5 M sodium acetate solution from inlet position “B1” was pumped slowly into the housing until all the air was removed and then the vent was closed The media was then equilibrated by pumping 24 mL of 0.5 M sodium acetate through the housing at a flow rate of 12 mL/min with the filtrate directed to outlet valve position “W” (waste) The ÄKTA system was then configured to deliver a 20 mM sodium acetate wash solution as a gradient comprising 4% of 0.5 M sodium acetate solution from inlet position “B1” and 96% deionized water from inlet position “A1.” 12 mL of the wash solution was pumped through the housing at a flow rate of mL/min with the filtrate directed to outlet valve position “W.” Then, the outlet valve position was switched to “Out1,” to direct the filtrate to the ISE assembly, and 12 mL of the 20 mM sodium acetate wash solution was pumped through the housing and ISE assembly at a flow rate of mL/min Finally, the injection valve was moved to change the inlet flow source to the 20 mM KCl challenge solution at sample pump position “S1,” and the challenge solution was pumped through the capsule at mL/min The potassium and chloride ISE responses were monitored, and flow was continued until complete breakthrough responses had been detected by both probes For each run, the filtrate volume at the initial inflection point of each of the potassium and chloride breakthrough responses was recorded as the corresponding breakthrough volume, and the AEX capacity was calculated as described in Section 2.4.1 2.5 Post-use installation validation bubble point test A post-use installation validation bubble point test was performed on capsules as specified by the manufacturer [26] Briefly, the capsule headspace was gravity drained and then the inlet was connected to a Sartocheck® Plus Filter Tester (Sartorius) while a length of tubing connected to the capsule outlet was immersed in water An upstream air pressure of 50 mbar was applied, and the air pressure was ramped up in 50 mbar increments until the bubble point pressure of the capsule was detected by visual obser5 J.F Hester, X Lu, J.D Calhoun et al Journal of Chromatography A 1654 (2021) 462445 vation of a stream of bubbles emerging from the outlet tubing A bubble point pressure greater than or equal to 900 mbar was considered a “pass,” while a bubble point pressure less than 900 mbar was considered a “fail” possibly indicating a mechanically defective capsule Note that, while some single-use AEX device manufacturers recommend similar gas diffusion type tests as pre-use tests [14,15], the bubble point test described above must be considered destructive for 3MTM Polisher ST and thus cannot be performed as a pre-use test, as it can introduce air between the FNW layers that may not be reliably removed relatively large capsule headspaces to enable rapid flow and handling of turbid fluids [12] The large resulting mixing volume upstream of the AEX media is responsible for the broad breakthrough curves relative to columns Selection of acetate as the standardizing anion is important; KCl challenges conducted on capsules standardized with sodium bicarbonate or sodium sulfate featured earlier and broader chloride breakthrough curves and reduced distinction between integral and damaged capsules (data not shown) This might be expected, since acetate has a lower selectivity coefficient with respect to the Q-functional media than chloride, resulting in a self-sharpening ion exchange boundary within the media, whereas more tightly-binding standardizing anions than chloride would be expected to result in an ion exchange boundary spreading with progression through the media [28] Attempts were made to calibrate the ISE’s such that the breakthrough curves could be plotted in units of ion concentration rather than as raw ISE electrical responses as in Fig These attempts were frustrated by significant ISE response drift over time originating from a number of factors Among these, both ISE’s are essentially measuring “zero” concentration prior to breakthrough of their respective analyte ions, the resulting cell potential of which is not well defined This accounts for the variable vertical positions of the breakthrough curves in Fig and other figures herein A second factor is relatively rapid drifts in the quantitative ISE responses over subsequent experiments, for example, due to plasticization of the polymeric membrane of the potassium ISE by acetate [29] Due to these challenges, ISE responses herein are plotted as raw electrical responses, since the primary response attributes of value are the filtrate volumes at which K+ and Cl− breakthrough occur 2.6 Controlled damage of AEX capsules To create various controlled levels of media damage, some 3MTM Polisher ST BC1, BC4, BC25, BC170, BC340, and BC1020 capsules were prepared with small, controlled defects by introducing holes through the full media stack using blunt-tipped needles (Small Hub RN Needle, Point Style 3, Hamilton, Reno, NV) of various diameter ranging from 32 gauge (0.11 mm) to 18 gauge (0.84 mm) In the case of BC1, BC4, and BC25 laboratory capsules, the holes were introduced in the stacked media prior to capsule assembly In the case of BC170 capsules, holes were introduced in the media in the internal lenticular filter element prior to assembling the capsule In the case of BC340 and BC1020 capsules, holes were introduced through one entire lenticular element, including media stacks on both sides of the element, prior to assembling the capsule One BC170 capsule was assembled using an undamaged lenticular filter element but with a purposeful misalignment of an o-ring that forms a seal between the filter element and the housing; this was expected to create a gross mechanical bypass within the capsule 3.2 Detection of missing media layers using pre-use AEX dynamic capacity test Results and discussion Fig is a plot of potassium and chloride ISE responses during the AEX dynamic capacity test for a complete media stack of layers of FNW and layers of FM in a 25 mm test housing (3 replicates), followed by ISE responses for sequential trials in which one layer of media was removed from the housing in each trial The ISE response inflection volumes were used to compute volumetric capacities as detailed in Table The full media stack exhibits a reproducible and passing (≥4 mL/cm2 ) volumetric chloride challenge capacity Media stacks with “missing” layers exhibit capacities that fail the test criterion and steadily decrease as media layers are removed While the chloride breakthrough volume shifts significantly with the removal of each media layer, the potassium breakthrough volume shifts only slightly, reflecting the smaller hold-up volume as the housing volume decreases with layer removal An empty housing exhibits nearly coincident potassium and chloride breakthrough events These results illustrate the capability of the non-destructive AEX dynamic capacity test to detect missing functional media layers and characterize the AEX capacity of single-use capsules prior to use, and also the hold-up volume independence of the method The potassium ISE response curves exhibit a second, minor inflection coincident with chloride breakthrough This is a result of the change in chloride concentration at the reference electrode liquid junction of the potassium ISE, the internal filling solution of which is aqueous sodium chloride [30] 3.1 Performance and repeatability of the pre-use AEX dynamic capacity test on integral capsules Five sequential AEX dynamic capacity tests were performed on each of three 3MTM Polisher ST BC340 capsules, produced using the same AEX media lots, as described in Section 2.4.1 One representative set of potassium and chloride responses for each capsule appears in Fig Potassium ISE responses featured a gradual, monotonically diminishing value prior to arrival of potassium ions at the ISE flow cell, followed by the rapid onset of an increasing ISE response which was recorded as the K+ breakthrough volume as illustrated in Fig The internal BC340 capsule volume of 730 mL [27] defines an expected lower limit of the K+ breakthrough volume In practice, measured potassium breakthrough volumes were somewhat larger and variable due to the volume of tubing between the challenge solution injection point and the ISE flow cell and also due to the fact that the test operator collected variable volumes of filtrate on the mass balance during the 20 mM sodium acetate equilibration step before switching the feed to the KCl challenge solution The chloride ISE responses featured a gradual, monotonically increasing value prior to chloride breakthrough, followed by the onset of a monotonically decreasing response which was recorded as the Cl− breakthrough volume as illustrated in Fig K+ and Cl− breakthrough volumes and calculated dynamic capacities for all 15 trials are detailed in Table The test was reproducible with a coefficient of variation well below 10 percent 3MTM Polisher ST capsules are designed for single-use AEX chromatography in a flow-through mode They feature shallow AEX bed depths relative to columns, and engineering trade-offs have been made to facilitate additional performance features, such as 3.3 Detection of damaged media using pre-use AEX dynamic capacity test Fig is a plot of potassium and chloride ISE responses during the AEX dynamic capacity test for an undamaged BC1 capsule and for BC1 capsules containing damaged media stacks, each J.F Hester, X Lu, J.D Calhoun et al Journal of Chromatography A 1654 (2021) 462445 Table ISE breakthrough volumes and calculated AEX dynamic binding capacities for 15 trials conducted on three 3MTM Polisher ST BC340 capsules produced using the same AEX media lots Capsule Capsule Capsule Capsule Capsule Capsule Capsule Capsule Capsule Capsule Trial K+ Breakthrough Volume, mL 794 859 868 756 837 Mean Standard Deviation Coefficient of Variation 848 908 1048 922 824 Mean Standard Deviation Coefficient of Variation 811 776 1917 786 902 Mean Standard Deviation Coefficient of Variation Cl− Breakthrough Volume, mL Volumetric capacity, mL/cm2 Cl− capacity, μequiv/cm2 2470 2597 2744 2433 2768 4.9 5.1 5.5 4.9 5.7 5.2 0.3 99 102 110 99 114 105 2802 2799 2956 2875 2705 5.7 5.5 5.6 5.7 5.5 5.6 0.1 2720 2711 3702 2632 2545 5.6 5.7 5.2 5.4 4.8 5.4 0.3 6.6% 115 110 112 115 111 113 2.0% 112 114 105 109 97 107 6.4% Fig Plot of chloride ISE response curves (top set of curves) and potassium ISE response curves (bottom set of curves) for 25 mm test housing containing various sets of AEX functional media layers Data collected using an ÄKTA avant 25 chromatography system J.F Hester, X Lu, J.D Calhoun et al Journal of Chromatography A 1654 (2021) 462445 Table ISE breakthrough volumes and calculated AEX dynamic binding capacities for 25 mm test housing containing various sets of AEX functional media layers Media Construction FNW + FM FNW + FM FNW + FM FNW + FM FM FM FM Empty housing Trial Trial Trial K+ Breakthrough Volume, mL Cl- Breakthrough Volume, mL Volumetric capacity, mL/cm2 Cl- capacity, μequiv/cm2 3.33 3.59 3.53 3.26 3.18 3.14 2.71 2.67 2.63 2.37 16.07 16.37 16.41 14.16 12.75 10.47 7.68 6.42 4.67 2.46 4.5 4.5 4.5 3.8 3.4 2.6 1.8 1.3 0.7 0.0 90 90 91 77 67 52 35 26 14 Fig Plot of chloride ISE response curves (top set of curves) and potassium ISE response curves (bottom set of curves) for BC1 (1 cm2 ) capsules containing media stacks pierced with blunt-tipped needles of various diameter All curves have been shifted on the x-axis such that the potassium breakthrough inflection point was set at zero volume Response curves have been intentionally vertically offset for easy comparison Y-axis tick marks are at intervals of 0.1 mA Hole sizes and net breakthrough volumes for each curve are indicated Data collected using the PendoTECH setup of which was pierced once with a blunt-tipped needle varying in diameter from 32 gauge (0.11 mm) to 22 gauge (0.41 mm) The media stack comprises two AEX media with different characteristics with respect to these piercing defects Whereas piercing of the FM creates a defined, roughly circular hole, the FNW swells upon wetting in buffer and appears capable of “healing” small puncture defects upon swelling At the largest hole size, it appears that a relatively well-defined low-pressure path is created in the media stack, and chloride breakthrough is observed to be nearly coincident with potassium breakthrough At smaller hole sizes, chloride breakthrough occurs earlier than for an undamaged capsule and the breakthrough curves broaden This is consistent with the superposition of fast chloride breakthrough within a low-pressure path comprising partially “healed” FNW layers and typical chloride breakthrough within the remainder of the media The pre-use AEX dynamic capacity test is quite sensitive to small defects, with a defect as small as 0.003% of the frontal media area having a volumetric capacity 4.84 LRV No virus plaques were observed in the plated viral challenge filtrate, however, and the value of 4.84 LRV was thus a lower limit of the viral clearance because the concentration of the spike in this particular challenge was insufficient to measure > LRV In all other of the 110 capsules tested, passage of the post-use bubble point test criterion was associated with ≥ LRV Phi-X174 clearance Conclusion Three efficiency tests have been developed and applied to characterize mechanical and chemical defects in six capsule formats of 3MTM Polisher ST, a single-use AEX chromatography product Two non-destructive, pre-use tests can be applied to reduce the risk of processing product-containing fluid with a damaged or defective capsule A pre-use pressure-based installation verification test is conveniently applied, with minimal equipment (only a pump and a pressure gauge or transducer), during the pre-conditioning flush procedure required to remove glycerin stabilizer from the capsule media and equilibrate the capsule for use This simple test is capable of detecting gross mechanical defects, such as broken internal seals, that might result from shipping damage, for example A new pre-use AEX dynamic capacity test is somewhat more complicated to execute, though it uses common reagents This test is more effective at reducing the risk of processing productcontaining fluid using a capsule containing even very small defects, with passing test results associated with > LRV viral clearance in a high-salt buffer challenge In addition, the AEX dynamic ca10 J.F Hester, X Lu, J.D Calhoun et al Journal of Chromatography A 1654 (2021) 462445 Fig Summary plot of observed post-use installation validation bubble point test values vs measured Phi-X174 viral clearance Upward arrows denote viral clearance experiments in which no virus plaques were observed after plating of the filtrate, and the minimal viral clearance was thus defined by the measured concentration of the challenge Vertical dashed line highlights the post-use bubble point test criterion Horizontal dashed line highlights a 5-LRV viral clearance level pacity test can detect chemical defects as well as mechanical ones and provides an in situ measurement of the capsule’s AEX capacity prior to running product-containing fluid A post-use installation validation bubble point test provides the greatest risk reduction with respect to mechanical defects, a passing test result corresponding to ≥ LRV viral clearance in a highsalt buffer challenge The bubble point test must be considered destructive for single-use devices like 3MTM Polisher ST that contain multiple layers of functional media Running it after capsule use, however, provides risk reduction with respect to any capsule damage that may have occurred either before or during use From the above three tests may be chosen a pair of pre-use and post-use efficiency tests that can be performed to substantially reduce the risk of viral clearance loss due to defects or damage in single-use AEX devices For operators concerned with minimizing risks associated with potential use of a single-use capsule to process product-containing fluid, an orthogonal combination of the pre-use AEX dynamic capacity test and the post-use bubble point test provides strong reduction of risk pre-use, as well as in situ characterization of device capacity Alternatively, a combination of the pre-use pressure-based installation verification test and the post-use bubble point test is very simple to perform and provides effective pre-use detection of the types of capsule defects expected to be the most frequent (e.g., gross bypass due to internal seal breakage resulting from shipping damage) Both combinations of tests provide an ultimate level of risk reduction associated with at least LRV viral clearance in a high-salt buffer challenge according to a data set based on 110 capsules tested in this study In either case, it is recommended that a fluid reprocessing option be included in regulatory filings for any AEX step, in case post-use validation were to detect the presence of a capsule defect Declaration of Competing Interest The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper CRediT authorship contribution statement Jonathan F Hester: Conceptualization, Methodology, Software, Validation, Formal analysis, Investigation, Writing – original draft, Writing – review & editing, Visualization Xinran Lu: Conceptualization, Methodology, Software, Validation, Formal analysis, Investigation, Data curation Jacob D Calhoun: Methodology, Validation, Formal analysis, Investigation, Data curation Rebecca A Hochstein: Methodology, Validation, Formal analysis, Investigation, Writing – original draft Eric J Olson: Methodology, Writing – review & editing References [1] , Viral safety evaluation of biotechnology products derived from cell lines of human or animal origin, in: Proceedings of the International Conference on Harmonisation of Technical Requirements for Registration of Pharmaceuticals for Human Use: Geneva, 1999 Switzerland [2] Center for Biologics Evaluation and Research Points to consider in the manufacture and testing of monoclonal antibody products for human Use U.S Food and Drug Administration: Rockville, MD, 1997; https://www.fda gov/regulatory-information/search-fda-guidance-documents/points-considermanufacture- and- testing- monoclonal- antibody- products- human- use [3] H.F Liu, J Ma, C Winter, R Bayer, Recovery and purification process development for monoclonal antibody production, MAbs (5) (2010) 480–499 [4] JC Giddings, Dynamics of Chromatography, Part 1: Principles and Theory, Marcel Dekker, Inc., 1965 11 J.F Hester, X Lu, J.D Calhoun et al Journal of Chromatography A 1654 (2021) 462445 [5] G.R Miesegaes, S.C Lute, E.K Read, KA Brorson, Viral clearance by flow-through mode ion exchange columns and membrane adsorbers, Biotechnol Prog 30 (1) (2014) 124–131 [6] L Norling, S Lute, R Emery, W Khuu, M Voisard, Y Xu, Q Chen, G Blank, K Brorson, Impact of multiple re-use of anion-exchange chromatography media on virus removal, J Chromatogr A 1069 (1) (2005) 79–89 [7] R.M O’Leary, D Feuerhelm, D Peers, Y Xu, GS Blank, Determining the useful life of chromatography resins: prospective small-scale studies, BioPharm 14 (9) (2001) 10–18 [8] B.D Kelley, J Jakubik, S Vicik, Viral clearance studies on new and used chromatography resins: critical review of a large dataset, Biologicals 36 (2) (2008) 88–98 [9] V Orr, L Zhong, M Moo-Young, C.P Chou, Recent advances in bioprocessing application of membrane chromatography, Biotechnol Adv 31 (4) (2013) 450–465 [10] T Vicente, M.F.Q Sousa, C Peixoto, J.P.B Mota, P.M Alves, M.J.T Carrondo, Anion-exchange membrane chromatography for purification of rotavirus-like particles, J Membr Sci 311 (1-2) (2008) 270–283 [11] J Weaver, S.M Husson, L Murphy, S.R Wickramasinghe, Anion exchange membrane adsorbers for flow-through polishing steps: part I Clearance of minute virus of mice, Biotechnol Bioeng 110 (2) (2013) 491–499 [12] J Hester, J Heitkamp, M Peters, Z Jokondo, J Rasmussen, Streamlined polishing and viral clearance using a new hybrid, biomimetic single-use anion exchanger, BioProcess Int 18 (10) (2020) 70–76 [13] J.X Zhou, F Solamo, T Hong, M Shearer, T Tressel, Viral clearance using disposable systems in monoclonal antibody commercial downstream processing, Biotechnol Bioeng 100 (3) (2008) 488–496 [14] Sartorius Stedim Biotech GmbH Operating instructions: Sartobind® Q and S Membrane Adsorbers, Sartorius order number 85037-549-25, 2015 [15] Pall CorporationMustang® Q, S and E Single-Use Capsules: Assembly and Installation Instructions, Pall Corporation Document Number USD, 2021 2933c https://biotech.pall.com/content/dam/pall/biopharm/lit- library/non- gated/ instructions- for- use/19- 10212- USD2933- c- Mustang- Q- S- E- Capsules- IFU- EN pdf?_ga=2.126645837.1598829374.1600449057-86485960.1600449057 [16] S Nochumson, P Levison, inventors; Pall Corporation, assignee System and method for testing chromatography media and devices U.S Patent 7,732,216 June 8, 2010 [17] F.M Plieva, I.N Savina, S Deraz, J Andersson, I.Y Galaev, B Mattiason, Characterization of supermacroporous monolithic polyacrylamide based matrices designed for chromatography of bioparticles, J Chromatogr B 807 (1) (2004) 129–137 [18] M.W Phillips, inventor; Millipore Corporation, assignee Method for determining an effective Peclet number for a membrane adsorber device U.S Patent 7,281,410 October 16, 2007 [19] W Demmer, R Faber, H.H Hoerl, A Thiefes, inventors; Sartorius Stedim Biotech GmbH, assignee Method for qualifying a non-particulate ionexchanger adsorber U.S Patent 8,969,094 March 3, 2015 [20] A Podgornik, J Vidic, J Jancar, N Lendero, V Frankovic, A Strancar, Noninvasive methods for characterization of large-volume monolithic chromatography columns, Chem Eng Technol 28 (11) (2005) 1435–1441 [21] N Lendero, J Vidic, P Brne, A Podgornik, A Strancar, Simple method for determining the amount of ion-exchange groups on chromatographic supports, J Chromatogr A 1065 (1) (2005) 29–38 [22] N Lendero, J Vidic, P Brne, V Frankovic, A Strancar, A Podgornik, Characterization of ion exchange stationary phases via pH transition profiles, J Chromatogr A 1185 (1) (2008) 59–70 [23] P.R Gavara, R Cabrera, R.R Vennapusa, M Grasselli, M Fernandez-Lahore, Preparation, characterization, and process performance of composite fibrous adsorbents as cation exchangers for high throughput and high capacity bioseparations, J Chromatogr B 903 (2012) 14–22 [24] J.F Hester, E.J Olson, F.E Porbeni, A.W Vail, A.M Voloshin, inventors; 3M Innovative Properties Company, assignee Method for testing a chromatography devices used for ion exchange WO Patent Application 2020/128797 June 25, 2020 [25] H Ruppach, Log10 reduction factors in viral clearance studies, BioProcess J 12 (4) (2014) 24–30 [26] 3M Purification Inc 2020 Installation qualification technical brief, 3MTM polisher ST series capsules, 3M document number 70-0216-0019-7 INS Rev 0920; https://multimedia.3m.com/mws/media/1878790O/3m- polisher- stinstallation- qualificaiton- tech- brief.pdf [27] 3M Purification Inc 2021 3MTM Polisher ST Regulatory Support File, 3M document number 10-698362 REV 0721 [28] F Helfferich, Theories of ion-exchange column performance: a critical study, Angew Chem Int Ed (8) (1962) 440–453 [29] U Oesch, W Simon, Life time of neutral carrier based ion-selective liquid-membrane electrodes, Anal Chem 52 (1980) 692–700 [30] F Bretschneider, J.R de Weille, in: Introduction to Electrophysiological Methods and Instrumentation, Academic Press, 2006, pp 103–131 12 ... Requirements for Registration of Pharmaceuticals for Human Use: Geneva, 1999 Switzerland [2] Center for Biologics Evaluation and Research Points to consider in the manufacture and testing of monoclonal... Calhoun et al Journal of Chromatography A 1654 (2021) 462445 Table Steps and settings for performing the pre-use integrity test on any capsule size using a PendoTECH control and data acquisition... Generally, Cytiva ÄKTA avant and ÄKTA pure chromatography system models are suitable for performing the pre-use AEX dynamic capacity test on 3MTM Polisher ST BC1, BC4, and BC25 capsules The polycarbonate