Increasing CHO cell harvest efficiency with cyclical cake filtration

Increasing CHO cell harvest efficiency with cyclical cake filtration | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Research Article Increasing CHO cell harvest efficiency with cyclical cake filtration Chloe Jerrold-Jones, Tizian Bucher, Andry Mannone, Abiram Gopalakrishnan, and 1 more This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-8605471/v1 This work is licensed under a CC BY 4.0 License Status: Posted Version 1 posted You are reading this latest preprint version Abstract Efficient clarification of bioreactor harvests remains a critical bottleneck in biomanufacturing, directly impacting throughput, product yield, and operational costs. Current industry practice typically relies on depth filters because they can effectively remove impurities, but they are also prone to fouling, require oversized filter areas, and can reduce product yield due to non-specific adsorption. As a response to some of those shortcomings, alluvial filtration was introduced which uses a diatomaceous earth (DE) cake to protect filter media, maintaining higher flow rates and extending filter life. Alluvial filters can reduce the required filter area, however, the true benefit of cake filtration can only be leveraged when using cyclical cake filtration. This technique uses multi-cycle cloth regeneration, achieving superior throughput and yield while maintaining filtrate clarity. Addition of ion exchange resins further allows for impurity reduction comparable to depth filtration without compromising product recovery. This study examines cyclical cake filtration and compares it with traditional and alluvial depth filtration regarding performance and scalability. Results indicate that cyclical cake filtration reduces required filter area and laboratory footprint, offering substantial economic and environmental benefits for commercial-scale production. CHO Cell Harvest Clarification Filtration Impurity Removal Scale-up Processing efficiency Figures Figure 1 Figure 1 Figure 2 Figure 3 Figure 3 Figure 4 Figure 5 Figure 6 Figure 7 Figure 8 Figure 9 Key Points Cyclical cake filtration offers an alternative to depth and alluvial filtration More efficient cell harvest at reduced filter area, lower waste and lower cost Higher product recovery was achieved with cyclical cake filtration Introduction With advancements in upstream design and the growing demand for pharmaceuticals and therapies, biomanufacturers are prioritizing efficient, high yield production processes (Auerbach 2025; Sharma et al. 2017; Thakur et al. 2020). After culture is grown in a bioreactor, clarification functions as the first step of removing cell debris, particulates, and ideally some impurities from the upstream harvest. Sometimes referred to as “midstream,” clarification acts as the gateway into downstream processing, known to be one of the biggest cost-intensive steps of bioproduction (Sharma et al. 2017; Rooij et al. 2019). Clarification is often a multi-stage operation, where primary clarification removes larger cell debris and secondary clarification removes finer impurities and particulates. For primary clarification, centrifuges and depth filters are popular options. For secondary clarification, typically another depth filter is used (Sharma et al. 2017). Disposable depth filters, although used widespread in industry – by 79% of respondents in the 16th Annual Report and Survey of Biopharmaceutical Manufacturing Capacity and Production – have various shortcomings especially with large scale manufacturing. Upstream variability or high cell densities can lead to dramatic decrease of flow rates and early fouling of depth filters (Minow et al. 2014). Cells commonly are captured directly within the tortuous path of the thick depth filter media (Fig. 1 a), forming a gel-like layer on the top. Nonuniform flow and pressure increases may trigger cells to break free and cause turbidity breakthroughs, jeopardizing reliable clarity requirements for filtrate (Nejatishahidein et al. 2022). Furthermore, once a depth filter membrane is clogged, the cartridge is considered spent and must be replaced. To ensure a complete batch can be processed without risk of early fouling, various sources in the literature suggest purposefully oversizing the filter area, referencing safety factors between 1.2 and 2.3 (Laska et al. 2005; Thakur et al. 2020; Dryden et al. 2021). Compensating this way may solve the problem of unreliable results when scaling up, but it comes at the cost of efficiency. One solution to improve the throughput of filters is to incorporate a filter aid, such as diatomaceous earth (DE). DE is made up of micron-scale diatom skeletons that have a unique incompressible and highly porous structure, and there are multiple commercially available types for GMP manufacturing that are basically inert (Daumke et al. 2019). Although adsorptive depth filters often include DE inside their cartridges, alluvial depth filters leverage DE by adding it directly to the feed stream (also known as body feed filtration) to form a filter cake on top of the media (Scott 2017; Meierrieks et al. 2023). Alluvial filters protect the filter media and dramatically reduce the risk of turbidity breakthrough, as the porous cake serves as the first line of defense to capture cell debris and particulates (Fig. 1 b). This additional filtration technique to remove process solids reduces the overall requirement of filter area, as the media is less prone to clogging. Some instances have shown a 3.5-fold increase in filter capacity when comparing alluvial filtration to centrifugation and depth filtration (Meierrieks et al. 2023). Filter capacity is directly correlated with economic and ecologic impacts, as less filter area is required, and thus, less waste generated. Despite these advantages, filter area remains a limiting factor to the scale-up of alluvial filtration, along with the volume of the cake (Ledergerber 2022; Daumke et al. 2019). Once a capsule is filled, it is considered waste and must be replaced with a new capsule. A recent advancement aimed at improving efficiency of cake filtration is the development of a single-use multi-cycle cake filter. This technology utilizes the same advantages of a porous filter cake to increase filter throughput and lifespan, with the added benefit of cloth regeneration step over multiple cycles (Fig. 1 c). Once the cake reaches a maximum volume, a sterile backflush step is automatically performed, cleaning the filter cloth and restoring its original flow rate. This is run in a cyclical function and can be repeated up to 50 times, massively increasing the capacity of the filter. Furthermore, the filter media is significantly thinner than depth filter media, allowing for higher flow rates, significantly reduced risk of fouling, and no requirement for preflushing. The selection of what technology to use for clarification is not only a nuanced challenge but can also have vast economic impacts on a process. Throughput, scalability, and material waste must be considered to improve overall efficiency of a harvesting step. This study investigates cyclical cake filtration and establishes a comparison to depth and alluvial filtration with respect to performance metrics, scalability, and practical implementation. Materials and Methods All experiments were conducted using an immunoglobulin G (IgG)-expressing ExpiCHO-S cell line (Gibco). The cultures that were used are summarized in Table 1 . For inoculum production and the subsequent batch and fed-batch cultivations, Efficient-Pro Medium (Gibco) supplemented with 6 mmol L − 1 L-glutamine and 0.1% Anti-Clumping Agent was used as basal medium. Efficient-Pro Feed 1 (Gibco) for cultivation 2 and Efficient-Pro Feed 2 (Gibco) for cultivation 3 were used throughout the fed-batch cultivations. Cells were cultivated in bioreactors with 5 to 10 L working volume. Table 1 ExpiCHO cell cultures used throughout the study. Culture 1 was used for Fig. 4 , culture 2 for Fig. 3 , and culture 3 for Figs. 5 – 9 Cultivation # Type Cell density [mio cells/ml] Viability [%] Crude Turbidity [NTU] Biomass [%] 1 Batch 9.5 99.2 454 1.0 2 Fed-batch 8.5 75.0 950 3.0 3 Fed-batch 10.7 96.3 870 2.8 The first two cultures were used for preliminary tests that were conducted once, while the third culture was used for the final comparison and scale-up where test conditions were repeated three times (with few exceptions as mentioned). All cultures were placed in the fridge after completing the cultivation and filtered on the subsequent day(s). For the cyclical cake filtration, a proprietary polypropylene filter cloth was used. All small-scale experiments were performed in a Nutsche that is shown in Fig. 2 a. A Nutsche consists of a cylindrical body with 200 ml volume underneath which the filter medium with 0.001 m 2 cross-sectional area is placed on a porous plate substrate. Pressurized air is applied from the top, and the pressure is set to a constant value via a regulator. The scale-up tests in Figs. 9 and 10 were performed with the FUNDALOOP® skid shown in Fig. 2 c and single-use bags having 0.006 m 2 and 0.018 m 2 cross sectional filter area (Fig. 2 b). For the alluvial filtration, cellulose and diatomaceous earth-based media with a retention range of 10–30 µm and 0.1–0.3 µm were used for the primary and secondary clarification step, respectively. The finer medium was charged with cationization agents. The tested filter area was 0.001 m 2 . For the depth filtration, cellulose and diatomaceous earth-based dual layer media with a retention range of 0.5–5 µm and 0.05–3 µm were used for the primary and secondary clarification step, respectively. The latter again had an anion exchange functionality to adsorb impurities. Lab filter capsules with a filter area of 0.0025 m 2 were used for the tests, with the exception of the test in Fig. 3 , where 0.001 m 2 was used. A Watson Marlow peristaltic pump with pump head 114DV and 3.2 mm inner diameter tubing was used to establish a constant flow, and a PendoTech single-use pressure sensor was used to measure the pressure. The pressure reading was monitored using a Levitronix LCO-i100 process control unit. To calculate the depth filter sizing in Fig. 8 , the P_max method was used (Merck KGaA 2020). The flow rates in all filtration methods were measured indirectly via a scale and taking periodic weight measurements, assuming that the density of the product is identical to the one of water. For the alluvial and cyclical cake filtration, diatomaceous earth (DE) was added to the feed prior to the filtration. For alluvial filtration, 65% per biomass Celpure C300 was added, which is within the typical range of 50–100% (van der Meer et al. 2014). This high-purity DE is specifically designed for biopharmaceutical applications and has a nominal permeability of 300 mDarcy. For the cyclical cake filtration, 65–100% per biomass was used. For Fig. 3 , 65% was used to have a meaningful comparison with alluvial filtration. Celpure C300 was again used for all experiments with the exception of one process condition tested in Figs. 5 – 7 where 80% per biomass Celpure S25E was used, having a nominal permeability of 25 mDarcy. The biomass was calculated from the total cell count and the average cell diameter, assuming a spherical shape of the cells. For cake washing, phosphate buffered saline (PBS) was used that was dissolved in water at a concentration of 9.55 g/L. The second trains of depth and alluvial filters (which are essentially normal depth filters due to the absence of DE) are typically charged in order to adsorb unwanted impurities such as deoxyribonucleic acid (DNA) and host cell proteins (HCP). In order to study the possibility of removing DNA and HCP, several ion exchange resins were compared in this study: INDION 810, a macroporous strongly basic type I anion exchange resin, INDION 820, a macroporous strongly basic type II anion exchange resin, INDION 830, a strong base type I anion exchange resin having quaternary ammonium function groups, and INDION 254, a strongly acidic cation exchange resin. Those ion exchange resins were contrasted with Daicel MabXPure that combines anion exchange (AEX) and size exclusion chromatography (SEC) and that has been featured in previous studies (Daumke et al. 2019, Daicel Bioseparations n.d.). The turbidity of the filtrate was measured using a WTW Turb 430 IR/T turbidity meter, and the cake thickness in cyclical cake filtration and alluvial filtration was measured using a caliper. To evaluate the filter performance, Immunoglobulin G (IgG), HCP, and DNA concentration were measured before and after filtration. Prior to the analysis the samples were centrifuged at 3000 g for 5 min to remove any remaining cells. IgG concentration was measured with the IgG Bio test kit (Roche Diagnostics) using the Cedex Bio Analyzer (Roche Diagnostics). For the DNA and HCP measurements, the samples were diluted according to the manufacturer’s recommendation. As the actual concentrations were unknown, three different dilutions were analyzed per sample. HCP concentration was measured using the CHO Host Cell Proteins ELISA Kit (RayBiotech). DNA concentration was measured using the Quant-iT Picogreen dsDNA Assay Kit (Invitrogen). Both HCP and DNA kits were evaluated using the Varioskan LUX spectrometer (Thermoscientific). Results In a Nutsche filter, 125 ml sample sizes of Culture 2 (Table 1 ) were filtered using a depth filter, alluvial filter, and cyclical cake filter at a constant pressure of 1.5 bar, which was chosen to limit cell damage (Minow et al. 2014). Depth filters are not normally run in a pressure-control mode (Sharma 2017), however this was done for the sake of understanding the filtration mechanisms. Complete blockage of the depth filter occurred after filtration of 125 ml. Figure 3 shows the flow rates in L/m 2 h (LMH) over time. As alluvial filtration and cyclical cake filtration exhibited a faster flow rate than depth filtration, data was extrapolated and displayed in respective dotted lines to show how the flow rate was expected to drop to an asymptotic plateau over time. Figure 4 compares the effect of resins on product yield and the removal of impurities. Yield was calculated by measuring the concentrations and volumes before and after the filtration. The volume before the filtration was corrected by the volume occupied by cells. In Fig. 4 a, five different ion-exchange resins were added to 160 ml samples of Culture 1 (Table 1 ) prior to filtration in the Nutsche with the cyclical cake filter cloth. Out of those tested, INDION 810 and MabXPure removed the most DNA and HCP and therefore were selected for further investigation in Figs. 4 b and 4 c. Figure 4 DNA, HCP, and IgG concentrations were compared with addition of various ion exchange resins. 100% refers to the initial concentration of DNA, HCP, and IgG in the feed. For all tests, 160 ml of Culture 1 and 100% Celpure C300 per biomass were filtered in the Nutsche with media from a cyclical cake filter. a) Five different ion-exchange resins were added to the feed. All INDION resins were added at ratios 1:2 relative to the weight of DE, and MabXPure was added at a ratio 1:5 to the feed. b) MabXPure was added in four volumetric ratios (1:50, 1:20, 1:10, 1:5) to the feed. c) INDION 810 was added in two ratios (1:5 and 1:2) to the weight of DE added, each for two mixing times (5 minutes or 30 minutes). All tests henceforth were done using Culture 3 (Table 1 ). For depth filtration, a capacity of 100 ml was determined via P_max method, and this filtration amount was used throughout the tests in Fig. 5 – 7 . For alluvial and cyclical cake filtration, 125 ml were filtered based on maximum cake thickness considerations. An exception was made for the condition using Celpure S25E and a finer filter cloth, where only 75 ml of culture was filtered. Depth and alluvial filtrations were divided into separate stages for primary and secondary clarification, in which filtrate from the first clarification was collected and measured before being fed into the second filter for further clarification. Both depth filters and alluvial filters were flushed with the prescribed washing liquid prior to filtration (135ml and 50ml, respectively). Alluvial filtration was performed using a peristaltic pump for the first test and then was switched to filtration in a Nutsche for the second and third test due to issues with DE sedimentation in the peristaltic pump tubing. Cyclical cake filters were not pre-flushed as per the operating instructions and were tested using four different process parameters in only one stage. Figure 5 Filtrate NTU was measured after processing Culture 3 with cyclical cake filtration in a single stage for four process conditions (always using 80% per biomass), and in two stages with alluvial filtration and depth filtration. Depth filters were run with a peristaltic pump at 10ml/min, or 240 LMH, and alluvial filters were run with a peristaltic pump at 15ml/min, or 900 LMH, as well as a Nutsche filter. Cyclical depth filters were processed in the Nutsche with a constant pressure of 1.5 bar. The following additions were tested: Celpure C300, Celpure S25E with a finer cloth, Celpure 300 with INDION 810 (1:5 ratio relative to DE), and Celpure C300 with MabXPure (1:10 ratio relative to feed). The error bars represent standard deviations. In Fig. 6 a, the same four testing conditions for cyclical cake filtration were compared to alluvial filtration and depth filtration in regard to product yield, both before and after washing with PBS. All tested scenarios showed a significant amount of product can be recovered with a washing step. Product yield was lowest after depth filtration and highest after cyclical cake filtration. The condition which resulted the highest product yield for cyclical cake filtration was with Celpure C300 alone. Figure 6 b compares the impurity levels of the filtrate. Depth filters showed the lowest level of both DNA and HCP. Alluvial filtration and cyclical depth filtration with the addition of MabXPure exhibited similar impurity profiles. The DNA level after filtration with Celpure C300 and S25E were higher than 100% because the filtration induces some amount of cell damage that gives rise to an increase in the DNA concentration. Figure 6 Product yield (based on IgG), DNA, and HCP levels from Culture 3 (Table 1) were measured using the same testing conditions as in Fig. 5. a) IgG levels were measured in the filtrate before and after a PBS wash. For cyclical cake filtration and alluvial filtration, 20% PBS wash per feed volume was used. For depth filtration, 25% PBS wash per feed volume was used. b) DNA and HCP levels were measured in the filtrate before washing. The error bars represent standard deviations. Figure 7 estimates the required filtration area to process 2’000 L of cell harvest for each filter type. Depth filtration required the most filter area, as was calculated using a P_max calculation, assuming a maximum duration of five hours processing time and a maximum pressure of 1.5 bar (Thakur et al. 2020; Cytiva 2023). Data shown for alluvial filtration reflects a maximum cake thickness of 18mm (Filtrox Group n.d.), and was extrapolated from earlier Nutsche trials, as cake thickness scales linearly with the filtered amount. Less filter area is required for alluvial filtration compared to depth filtration. For cyclical cake filtration, it was determined from previous Nutsche tests that 100 ml per filtration with a 0.001m 2 filter area yielded a cake of approximately 9.5 mm, allowing for a successful discharge. It is assumed that the first quarter of the filtrate is recycled back as prefiltrate, reducing the amount that can be filtered with 0.001m 2 to 75 ml. Using these values to scale up to two commercial-sized FUNDALOOP® filters with a combined filter area of 0.88m 2 , it is calculated that 66 L can be filtered per cycle, allowing for the filtration of 2’000 L in 30 cycles. Figure 7 The required filter area to process 2,000 L of CHO culture was calculated for the three filter methods. It was assumed that secondary clarification for alluvial and depth filtration requires half of the filter area as primary clarification (Zustiak et al. 2022; Merck KGaA 2017). A typical safety factor of 50% was included for alluvial and depth filtration (Dryden et al. 2021). Figure 8 compares flow rates between three scales for cyclical cake filtration: the Nutsche, FUNDALOOP® 50, and FUNDALOOP® 200. The tested volumes at the three scales were 125 ml, 750 ml, and 2,250 ml, respectively. With the exception of a brief initial surge in flow recorded in the Nutsche, the flow rates exhibited similar patterns and values between scales. Figure 8 Flow rates in L/m2h were measured over time in three scales of cyclical cake filtration with Culture 3 (Table 1). The filter areas tested were 0.001m2 (Nutsche), 0.006m2 (FUNDALOOP® 50), and 0.018m2 (FUNDALOOP® 200). 80% Celpure C300 was added per biomass, as well as INDION 810 at a 1:5 ratio relative to DE. The error bars represent standard deviations. Filtrate quality parameters were compared between the same three scales for cyclical cake filtration in Fig. 9 . Product yield was consistent and similar across all scales. DNA and HCP were less consistent than yield, but within range of each other, especially considering the larger standard deviations associated with the measurement. Turbidity was lower in the Nutsche than the FUNDALOOP® 50 or 200. Figure 9 Turbidity, product yield (based on IgG), DNA, and HCP levels were compared over three scales of cyclical cake filtration with Culture 3 (Table 1). Turbidity was measured in NTU and the remaining parameters were measured in a relative concentration to the feed material by the same methodology as described in Fig. 4. Celpure C300 and INDION 810 were added as in Fig. 8. The FUNDALOOP® 200 was only tested once for turbidity and yield; all other tests were conducted three times. The error bars represent standard deviations. Discussion A critical element of efficiency in bioproduction involves how quickly a batch can be processed. Faster flow rates during cell harvesting and clarification is one way to decrease processing time and increase efficiency. As shown in Fig. 3 , flow rates with depth filter media slowed down quickly and approached an asymptotic plateau faster than alluvial filtration or cyclical cake filtration. This can be explained by the mechanism of capture; cells and debris are retained by the thick depth filter media itself, which leads to rapid clogging. After filtering 125mls of Culture 1, the media was almost completely fouled. In contrast, no clogging occurred in alluvial filtration, as the cells and debris are chiefly held back by the filter cake, which protects the filter media and maintains a higher flow rate. Had the filtration been continued, the flow rate would have also continued to drop due to the increased resistance of the cake, but even in the asymptotic plateau the flow rates would have been considerable. Unlike in depth filtration, the capacity is defined either by the maximum cake thickness – a physical limitation of the filter – or a specified minimum flow rate. The mechanism of capture in cyclical cake filtration is similar to alluvial filtration, but the flow rates are even higher due to the thin filter media having a lower resistance than the thick and tortuous depth filter media. The maximum capacity per cycle could be even higher than in alluvial filtration, however, the capacity is usually limited by the maximum allowable cake thickness – typically between 10-12mm. Nonetheless, the overall capacity is extensively increased with the allowance of 50 cycles to be performed using one cloth. Overall, Fig. 3 demonstrates that cyclical cake filtration significantly decreases the process risk during the primary clarification step. The incorporation of filter aid eliminates the risk of clogging – one of the key downfalls of depth filters that often leads to a partial loss of product titers. Additionally, there is no risk of overfilling filters, as is a concern in alluvial systems, because a more challenging filtration is tackled with an increased number of filtration cycles. Variabilities in the production or feed streams can therefore be effectively handled without the need of adding more filter area. While this study shows that depth filters struggled to achieve high throughput and would require an exorbitant amount of filter area at scale, they are known to capture a large portion of impurities with isoelectric bonds by charging their membrane. However, this advantage could be rivaled in alternative filtration methods by the addition of ion exchange resins to the feed. Figure 4 b and 4 c show two resin options for reduction of impurities in cyclical cake filtration: MabXPure and INDION 810. The addition of MabXpure at a 1:5 ratio of resin to feed reduced the DNA and HCP levels the most, however this did come at an expense to the product yield. Reducing the ratio of MabXPure to 1:10 increased the product yield to approximately 90% while also removing a majority of the impurities (Fig. 4 b). Furthermore, INDION 810 resulted in a higher product yield of approximately 95%, and still removed the majority of DNA and HCP even with a low ratio, as long as the resin was incubated for 30 minutes prior to filtration. IgG levels were unaffected by INDION 810 ratio or incubation time. While a reduced impurity profile helps to reduce the burden of further downstream steps, other characteristics of high filtrate quality include filtrate clarity and product yield. These parameters were compared between cyclical cake filtration, alluvial filtration, and depth filtration in Figs. 5 and 6 . It is important to note that alluvial filtration and depth filtration were each run in a two-stage filter train, standardly referred to as primary and secondary clarification. The data shown from cyclical cake filtration is obtained from only one round of filtration. For a standard filtration where 80% Celpure C300 per biomass was used, the turbidity of cyclical cake filtration roughly matched the turbidity achieved after the first clarification with alluvial and depth filters. With the addition of MabXPure in a 1:10 ratio of resin to feed and 80% Celpure C300 per biomass, the turbidity could be significantly reduced to less than 5 NTU and rivaled the filtrate clarity of depth filtration after primary and secondary clarification (Fig. 5 ). The possibility to achieve similar clarity levels in one step versus two could have extreme impacts on cost, time, and complexity of cell harvesting in a bioprocess. Another cyclical cake condition shown in Fig. 5 that reduced filtrate NTU was using a finer DE grade (Celpure S25E) and a finer cloth. Compared to Celpure C300, the smaller size and lower permeability of Celpure S25E aids in capturing smaller micron sized particles. This DE substitution yielded a filtrate NTU of approximately 12 in a single step, similar to the level which was reached after the second stage of alluvial clarification. This trade-off reflects a well-documented phenomenon in filtration systems, where tighter filter media typically yield improved filtrate clarity at the expense of increased pressure buildup and reduced flow rate. These parameters are interdependent and therefore must be carefully balanced when optimizing filtration performance. In addition to reducing the filtrate NTU, MabXPure also significantly improved the impurity profile of the filtrate (Fig. 6 b). The levels of DNA and HCP in the filtrate from cyclical cake filtration with the addition of MabXPure were nearly identical to that of alluvial filtration. Although depth filtration still resulted in the lowest impurity profile, the addition of MabXPure provides an alternative option to reduce DNA and HCP without depth filtration. It is prudent to note that data shown in Fig. 6 b was measured pre-washing with PBS, and future studies must be done to measure the impurity levels post washing to understand if this would increase DNA or HCP levels. It is likely that the addition of ion exchange resins in the cake would tightly bond impurities and therefore retain them well even during washing, but this needs to be investigated further. It also needs to be noted that due to the aforementioned challenges with DE sedimentation in the peristaltic pump tubing, the second and third alluvial filtrations were carried out in a Nutsche. These tests were therefore subjected to higher flow rates than are suggested by the manufacturer due to operational limitations of the Nutsche. It is probable that if alluvial filtration had been conducted with slower flow rates, longer contact time with the filter media would have resulted in more effective capture of DNA and HCP, similar to the levels shown with depth filtration. This would be best confirmed in a future study. Although the surface charge on depth filter media can aid in impurity removal, it is non-specific and can therefore capture any molecule that it has an electrostatic interaction with. This can include the product of interest, in turn resulting in a lower product yield. This is exemplified in Fig. 6 a, with depth filters having the lowest product yield compared to alluvial and cake filtration. Product can also be retained by the filter cake and cloth in cyclical cake filtration and alluvial filtration, but the effect of this is less pronounced. The highest product yield was observed after cyclical cake filtration with the addition of Celpure C300 without ion exchange resin. The finer DE grade and additional resins resulted in a slightly lower product yield, albeit higher than depth filtration and alluvial filtration. When considering all results of Figs. 5 and 6 , cyclical cake filtration with the addition of Celpure C300 and MabXPure was able to achieve very similar results as two-stage depth or alluvial filters in terms of impurity removal and filtrate clarity, albeit at a higher product yield. These results were obtained using IgG as the model product. Because different molecules exhibit distinct isoelectric points (pI) and adsorptive properties, product yield may vary substantially depending on the process conditions. Therefore, thorough optimization is necessary to determine which filter aids and additives, if any, are most suitable for each product and process. Furthermore, although these findings demonstrate that the addition of MabXPure can reduce impurity levels in the filtrate while achieving a high product yield, the cost of resin additives could offset the benefits. Harvest conditions and requirements should be evaluated individually to determine if resin addition is a valuable and economic option. It can be argued that filtrate quality is the most critical measure of filtration performance, but practical implementation will often depend on a proper scale-up evaluation. In Fig. 7 , data was extrapolated from small scale trials to estimate required filtration area for a 2,000 L batch, commonly used as a commercial scale production volume when using single-use bioprocessing equipment. Alluvial filtration significantly reduces filter area requirement compared to depth filters, as the filter cake extends the life of the media. Further supporting this, Minow et al. (2014) reported that scaling up depth filters to commercial levels often times does not make financial sense due to low filtration capacities per filter area. However, both depth filters and alluvial filters are typically oversized and installed with a safety factor. In alluvial filters, this safety factor is intended to avoid filter filling, whereas in depth filters it primarily prevents filter blockage. Cyclical cake filtration leverages the benefit of a protective cake layer in alluvial filtration even further with its backflush step and regenerative nature, causing the required surface area to dramatically drop. Moreover, there is no need for an additional safety factor, as each filter can withstand up to 50 cycles. The 30 cycles calculated to process 2,000 L in Fig. 7 would result in 66% safety factor over the maximum allowable number of cycles. Overall, cyclical cake filtration allows for 98% and 96.5% filter area savings compared to depth and alluvial filtration, respectively. Required filtration area directly correlates to operational cost, potential for extractables and leachables to compromise a batch, and quantity of plastics and single-use materials that are discarded after each batch. Furthermore, capital equipment requirement and therefore cleanroom space also increase with higher consumable usage. Implementation of cyclical cake filtration as a harvest method can have substantial economic and environmental benefits, creating a leaner and more efficient bioprocess step. It needs to be mentioned that this study did not conduct any optimization on the depth and alluvial filtrations and simply compared models that are standard in the industry for both. In recent years, those competing technologies have been continuously developed further by manufacturers, resulting in products on the market that achieve better results in terms of capacity, yield, and removal of impurities. However, those filters still have the same drawbacks (can only be used once until clogging, are not regenerable, and require a large filter area), and it is expected that general trends in the data remain the same. Scalability between individual cyclical cake filtration systems is shown in Figs. 8 and 9 , detailing flow rates and filtrate quality. Between the Nutsche system and FUNDALOOP® 200, there is an 18-fold increase in the filter area. Flow rates across all three systems behave similarly, apart from an initial high flow rate in the Nutsche as the operational mechanism differs and 1.5 bar is achieved instantly. In the FUNDALOOP® systems, pressure builds up slower to account for bag filling and formation of the cake, resulting in a lower initial flow rate. After pressure is established, flow patterns between the three systems are very similar. The quality profile of filtrate was comparable between the three scales as well. There were slight variations in impurity levels, but product yield was relatively unaffected by the difference in scales. The turbidity was lower in the Nutsche system, however, which can be explained by the absence of turbulence inside the filter vessel. In the FUNDALOOP® system, upwards flow of feed creates an added element of homogenization and filtration is not working in the direction of gravity. The observed variation of filtrate clarity is likely a function of this mechanistic difference as opposed to a change to the filtration area. Within the different FUNDALOOP® sizes, the turbidity values are stable and consistent. A larger system, the FUNDALOOP® 3000, is also commercially available and exhibits approximately a 24-fold increase compared to the FUNDALOOP® 200. While tests with the FUNDALOOP® 3000 were outside the scope of this study, the presented data likely supports the scalability of this system as well. Conclusion This study demonstrates that cyclical cake filtration delivers a highly efficient and scalable alternative to depth and alluvial filtration for conventional CHO harvest clarification, providing faster flow rates, higher product yield, reduced fouling, and significantly lower filtration area requirements. The use of filter aids protects the filter medium, eliminating the danger of clogging that is a key disadvantage of depth filters. The thin filter medium and regenerative cycling improve throughput, while the incorporation of ion exchange resins such as MabXPure or INDION 810 enables substantial removal of DNA and HCP – achieving filtrate clarity and impurity levels comparable to multi-stage depth or alluvial filtration in a single unit operation. Additionally, performance can be tailored by adjusting DE grade, cake thickness, pressure limit, resin type, resin ratio, and incubation time, offering process flexibility for optimization. Collectively, these results highlight cyclical cake filtration as a robust and cost-effective technology to enhance midstream bioprocess efficiency. Author Contribution Statement CJJ performed part of the filtration tests, helped with data analysis and interpretation, wrote the majority of the manuscript, and compiled references. TB designed the study, executed the filtration tests, performed the data analysis, and helped writing the manuscript. AM and AG did the cultivations and executed the IgG, DNA, HCP measurements and analyses. JO coordinated the efforts from the ZHAW and helped with the data analyses and interpretation. All authors read and approved the manuscript. Declarations Competing Interests Financial interests: Author CJJ and TB are employees of DrM, which manufactures the filtration system evaluated in this study. Non-financial interests: The authors have no non- financial competing interests to declare. Funding No funding was received to assist with the preparation of this manuscript. Author Contribution CJJ performed part of the filtration tests, helped with data analysis and interpretation, wrote the majority of the manuscript, and compiled references. TB designed the study, executed the filtration tests, performed the data analysis, and helped writing the manuscript. AM and AG did the cultivations and executed the IgG, DNA, HCP measurements and analyses. JO coordinated the efforts from the ZHAW and helped with the data analyses and interpretation. All authors read and approved the manuscript. Acknowledgement We would like to thank Samuel Schneider for his support during the HCP and DNA analysis, as well as Dardan Qereti for the support during experiments. References th Annual Report and Survey of Biopharmaceutical Manufacturing Capacity and Production (2019) BioPlan Associates, Inc. www.bioplanassociates.com/16th Accessed 12 Nov 2025 Auerbach M (2025) The bioprocess revolution: how technology and trends are reshaping pharmaceutical manufacturing. American Pharm Review. https://www.americanpharmaceuticalreview.com/Featured-Articles/618316-The-Bioprocess-Revolution-How-Technology-and-Trends-are-Reshaping-Pharmaceutical-Manufacturing / Accessed 22 Oct 2025 Cytiva (2023) Stax™ mAx Clarification Platform. https://cdn.cytivalifesciences.com/api/public/content/kfGqaqgYvEi8m_T7wAWa1A-pdf Accessed 12 Nov 2025 Daicel Bioseparations (n d.) Static performance of MabXPure: application note #1. https://www.daicelchiral.com/wp-content/themes/daicel/assets/add_img/pdf/DAICEL-Bioseparations_Static-Performance-of-MabXpure-Application-Note-1.pdf Accessed 17 Dec 2025 Daumke R, Medvedev V, Albano T, Rousset F (2019) Alluvial filtration: an effective and economical solution for midstream application. In: Eibl R, Eibl D (eds) Single-Use Technology in biopharmaceutical manufacture, 2nd edn. Wiley, Inc, pp 271–277 Dryden W, Larsen L, Britt D, Smith M (2021) Technical and economic considerations of cell culture harvest and clarification technologies. Biochem Eng J 167:107892. doi.org/10.1016/j.bej.2020.107892 Filtrox Group (n d.) PURADISC® SU technical data sheet. https://www.filtrox.com/applications/filtration-in-life-science/single-use/ Accessed 16 Dec 2025 Laska M, Brooks R, Gayton M, Pujar N (2005) Robust scale-up of dead end filtration: impact of filter fouling mechanisms and flow distribution. Biotechnol Bioeng 92(3):308–320. 10.1002/bit.20587 Ledergerber B (2022) Upscaling: scaling of a filtration process. Filtrox AG. https://www.filtrox.com/wp-content/uploads/2022/08/AN_LS_005_Upscale_Jul2022_EN_note.pdf Accessed 12 Oct 2025 Meierrieks F, Pickl A, Wolff M (2023) A robust and efficient alluvial filtration method for the clarification of adeno-associated viruses from crude cell lysates. J Biotechnol 367:31–41. doi.org/10.1016/j.jbiotec.2023.03.010 Merck KGA (2017) Clarification of mammalian cell cultures by depth filtration. https://www.sigmaaldrich.com/deepweb/assets/sigmaaldrich/product/documents/350/422/pf1804en-mk.pdf Accessed 16 Dec 2025 Merck KGA (2020) Pmax™/Tmax™ constant flow rate test for depth filter sizing. https://www.sigmaaldrich.com/deepweb/assets/sigmaaldrich/product/documents/538/353/pmax-tmax-protocol-br4506en-mk.pdf , accessed 16 Dec 2025 Minow B, Egner F, Jonas F, Lagrange B (2014) High–cell-density clarification by single-use diatomaceous earth filtration (2025) Bioprocess Int. https://www.bioprocessintl.com/single-use/high-cell-density-clarification-by-single-use-diatomaceous-earth-filtration Accessed 12 Oct Nejatishahidein N, Kim M, Jung S, Borujeni E, Fernandez-Cerezo L, Roush D, Borhan A, Zydney A (2022) Scale‐up issues for commercial depth filters in bioprocessing. Biotechnol Bioeng 119:1105–1114. 10.1002/bit.28035 Rooij J, DeConto J, Schaenzler G, Bauer D, Barre K, Duskin M, Kohli A, Watanabe K (2019) Upstream and downstream solutions for AAV manufacturing. Cell Gene Therapy Insights 5(Suppl 5):1017–1029. 10.18609/cgti.2019.110 Scott C (2017) Downstream disposables: the latest single-use solutions for downstream processing. BioProcess Int 15(1)i:1–5 Sharma M, Raikar S, Srivastava S, Gupta S (2017) Examining single-use harvest clarification options: a case study comparing depth-filter turbidities and recoveries. BioProcess Int 15(2):40–47 Thakur G, Hebbi V, Parida S, Rathore A (2020) Automation of dead end filtration: an enabler for continuous processing of biotherapeutics. Front Bioeng Biotechnol 8:758. 10.3389/fbioe.2020.00758 van der Meer T, Minow B, Lagrange B, Krumbein F, Rolin F (2014) Diatomaceous Bioprocess Int 2(8):S25-S28 earth filtration: innovative single-use concepts for clarification of high-density mammalian cell cultures, Zustiak M, Smith M, Luo Q, Kruger J, Dryden W, Rivera M (2014) (2022) Simple, scalable single-stage harvest solution using chromatographic clarification technologies. Thermo Fisher Scientific Inc. https://documents.thermofisher.com/TFS-Assets%2FBPD%2FReference-Materials%2Fwhite-paper-single-use-harvest-solutions.pdf Accessed 16 Dec 2025 Additional Declarations Competing interest reported. Financial interests: Author CJJ and TB are employees of DrM, which manufactures the filtration system evaluated in this study. Non-financial interests: The authors have no non- financial competing interests to declare. Cite Share Download PDF Status: Posted Version 1 posted You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-8605471","acceptedTermsAndConditions":true,"allowDirectSubmit":true,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":581039184,"identity":"4667d47c-a838-4d76-b260-1d137ece510f","order_by":0,"name":"Chloe Jerrold-Jones","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAA8UlEQVRIiWNgGAWjYDACCeYGIMkGYjI+4CFOCyNcC7MBKVrAgE2CKC38sxsbH3xg4JMzZ29/VvF2zza77e09Zg8YamyicVpy52Cz4QwGNmPLnjNmN+c8u50858wZcwOGY2m5DTi0GEgktknzMLAlbriRw3ab58DtZAmJHDOgaw/j1/IHpOX+82fFxGthANvCYMYM1GJHUIvEjcRmwx4DNmODMznGknMO3E6Q4DlWbpCAxy/8M5IPPvhRcUzO4Pjxhx/eHLhtL8HevO3BhxobnFqgzjsGZyY2gCI2Aa9yMKiBs+wZoGlhFIyCUTAKRgEMAABZi1kRtTqFEQAAAABJRU5ErkJggg==","orcid":"","institution":"DrM inc","correspondingAuthor":true,"prefix":"","firstName":"Chloe","middleName":"","lastName":"Jerrold-Jones","suffix":""},{"id":581039185,"identity":"5fd72c09-54e1-4cf2-b805-3f60bc3df079","order_by":1,"name":"Tizian Bucher","email":"","orcid":"","institution":"DrM Dr. Mueller AG","correspondingAuthor":false,"prefix":"","firstName":"Tizian","middleName":"","lastName":"Bucher","suffix":""},{"id":581039186,"identity":"14d9d01d-f3da-43e6-9272-25cae2d76ce0","order_by":2,"name":"Andry Mannone","email":"","orcid":"","institution":"Zürich University of Applied Sciences","correspondingAuthor":false,"prefix":"","firstName":"Andry","middleName":"","lastName":"Mannone","suffix":""},{"id":581039187,"identity":"fb7fd0eb-6c4c-427c-be3a-6387ce960af3","order_by":3,"name":"Abiram Gopalakrishnan","email":"","orcid":"","institution":"Zürich University of Applied Sciences","correspondingAuthor":false,"prefix":"","firstName":"Abiram","middleName":"","lastName":"Gopalakrishnan","suffix":""},{"id":581039188,"identity":"7e8844cc-5619-420a-951c-f6a9f7912922","order_by":4,"name":"Jan Ott","email":"","orcid":"","institution":"Zürich University of Applied Sciences","correspondingAuthor":false,"prefix":"","firstName":"Jan","middleName":"","lastName":"Ott","suffix":""}],"badges":[],"createdAt":"2026-01-14 23:23:17","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-8605471/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-8605471/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":101751422,"identity":"1d24348d-ebb7-4983-a72b-84696b1f4e5f","added_by":"auto","created_at":"2026-02-03 10:20:09","extension":"jpg","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":701655,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cem\u003eIllustrations of how filters behave and actual photos of filter cloths and cake for a) depth filters, b) alluvial filters, and c) cyclical cake filters.\u003c/em\u003e\u003c/p\u003e","description":"","filename":"Fig.1.jpg","url":"https://assets-eu.researchsquare.com/files/rs-8605471/v1/761eef8dc9601184d9ebcf74.jpg"},{"id":101516643,"identity":"fed8173d-737b-403d-afde-f0b0f77308c6","added_by":"auto","created_at":"2026-01-30 16:13:14","extension":"jpg","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":701655,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cem\u003eIllustrations of how filters behave and actual photos of filter cloths and cake for a) depth filters, b) alluvial filters, and c) cyclical cake filters.\u003c/em\u003e\u003c/p\u003e","description":"","filename":"Fig.1.jpg","url":"https://assets-eu.researchsquare.com/files/rs-8605471/v1/dd1babd8f3dd773d143a2cee.jpg"},{"id":101420891,"identity":"4e3deb79-3de4-4055-aa35-b58208099c05","added_by":"auto","created_at":"2026-01-29 13:29:07","extension":"jpg","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":1213302,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cem\u003ea) Photo of stainless steel 200 ml Nutsche filter; b) three scales of the FUNDALOOP® consumable filters: 0.001m\u003c/em\u003e\u003csup\u003e\u003cem\u003e2\u003c/em\u003e\u003c/sup\u003e\u003cem\u003e (left), 0.006m\u003c/em\u003e\u003csup\u003e\u003cem\u003e2\u003c/em\u003e\u003c/sup\u003e\u003cem\u003e (middle), and 0.018m\u003c/em\u003e\u003csup\u003e\u003cem\u003e2\u003c/em\u003e\u003c/sup\u003e\u003cem\u003e (right); c) fully automatic FUNDALOOP® 50/200 skid\u003c/em\u003e\u003c/p\u003e","description":"","filename":"Fig.2.jpg","url":"https://assets-eu.researchsquare.com/files/rs-8605471/v1/a45b9f9f3c97192fa7fab69f.jpg"},{"id":101751540,"identity":"6b892111-2a26-4942-a58b-a630b3253e17","added_by":"auto","created_at":"2026-02-03 10:21:10","extension":"jpg","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":205128,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cem\u003e125 ml culture was filtered using a depth, alluvial, and cyclical cake filter at a constant pressure of 1.5 bar. 65% DE per biomass was used for the alluvial and cyclical cake filtration, and the data was extrapolated for comparison since those filtrations terminated faster than depth filtration. All tests were performed one time.\u003c/em\u003e\u003c/p\u003e","description":"","filename":"Fig.3.jpg","url":"https://assets-eu.researchsquare.com/files/rs-8605471/v1/c137630f10c5c46c78c9c668.jpg"},{"id":101516723,"identity":"6fdc9318-63f8-4ede-9710-1913dd8b6383","added_by":"auto","created_at":"2026-01-30 16:15:28","extension":"jpg","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":205128,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cem\u003e125 ml culture was filtered using a depth, alluvial, and cyclical cake filter at a constant pressure of 1.5 bar. 65% DE per biomass was used for the alluvial and cyclical cake filtration, and the data was extrapolated for comparison since those filtrations terminated faster than depth filtration. All tests were performed one time.\u003c/em\u003e\u003c/p\u003e","description":"","filename":"Fig.3.jpg","url":"https://assets-eu.researchsquare.com/files/rs-8605471/v1/e715f692d79addc29f5f4df2.jpg"},{"id":101420890,"identity":"f4b2000f-44b9-4c0b-b2b4-c6e0a8d52ace","added_by":"auto","created_at":"2026-01-29 13:29:07","extension":"png","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":282323,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cem\u003eDNA, HCP, and IgG concentrations were compared with addition of various ion exchange resins. 100% refers to the initial concentration of DNA, HCP, and IgG in the feed. For all tests, 160 ml of Culture 1 and 100% Celpure C300 per biomass were filtered in the Nutsche with media from a cyclical cake filter. a) Five different ion-exchange resins were added to the feed. All INDION resins were added at ratios 1:2 relative to the weight of DE, and MabXPure was added at a ratio 1:5 to the feed. b) MabXPure was added in four volumetric ratios (1:50, 1:20, 1:10, 1:5) to the feed. c) INDION 810 was added in two ratios (1:5 and 1:2) to the weight of DE added, each for two mixing times (5 minutes or 30 minutes).\u003c/em\u003e\u003c/p\u003e","description":"","filename":"4.png","url":"https://assets-eu.researchsquare.com/files/rs-8605471/v1/16560356fc3f0caeccc74190.png"},{"id":101420892,"identity":"5df96fb1-e33a-48a1-92ea-e85cc43ef66e","added_by":"auto","created_at":"2026-01-29 13:29:07","extension":"jpg","order_by":5,"title":"Figure 5","display":"","copyAsset":false,"role":"figure","size":145939,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cem\u003eFiltrate NTU was measured after processing Culture 3 with cyclical cake filtration in a single stage for four process conditions (always using 80% per biomass), and in two stages with alluvial filtration and depth filtration. Depth filters were run with a peristaltic pump at 10ml/min, or 240 LMH, and alluvial filters were run with a peristaltic pump at 15ml/min, or 900 LMH, as well as a Nutsche filter. Cyclical depth filters were processed in the Nutsche with a constant pressure of 1.5 bar. The following additions were tested: Celpure C300, Celpure S25E with a finer cloth, Celpure 300 with INDION 810 (1:5 ratio relative to DE), and Celpure C300 with MabXPure (1:10 ratio relative to feed). The error bars represent standard deviations.\u003c/em\u003e\u003c/p\u003e","description":"","filename":"Fig.5.jpg","url":"https://assets-eu.researchsquare.com/files/rs-8605471/v1/3d69e8b23c72f0f740715859.jpg"},{"id":101420886,"identity":"7b4e8ec1-2d6c-4ccd-abda-4d117bd3bb3a","added_by":"auto","created_at":"2026-01-29 13:29:06","extension":"png","order_by":6,"title":"Figure 6","display":"","copyAsset":false,"role":"figure","size":163443,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cem\u003eProduct yield (based on IgG), DNA, and HCP levels from Culture 3 (Table 1) were measured using the same testing conditions as in Fig. 5. a) IgG levels were measured in the filtrate before and after a PBS wash. For cyclical cake filtration and alluvial filtration, 20% PBS wash per feed volume was used. For depth filtration, 25% PBS wash per feed volume was used. b) DNA and HCP levels were measured in the filtrate before washing. The error bars represent standard deviations.\u003c/em\u003e\u003c/p\u003e","description":"","filename":"6.png","url":"https://assets-eu.researchsquare.com/files/rs-8605471/v1/43b146e4a44c4b77d0cbbe82.png"},{"id":101420893,"identity":"14ba23c1-4a05-4e21-8555-527015d58157","added_by":"auto","created_at":"2026-01-29 13:29:07","extension":"jpg","order_by":7,"title":"Figure 7","display":"","copyAsset":false,"role":"figure","size":241815,"visible":true,"origin":"","legend":"\u003cp\u003eThe required filter area to process 2,000 L of CHO culture was calculated for the three filter methods. It was assumed that secondary clarification for alluvial and depth filtration requires half of the filter area as primary clarification (Zustiak et al. 2022; Merck KGaA 2017). A typical safety factor of 50% was included for alluvial and depth filtration (Dryden et al. 2021).\u003c/p\u003e","description":"","filename":"Fig.7.jpg","url":"https://assets-eu.researchsquare.com/files/rs-8605471/v1/77a2597c358e1b7a1f5cd202.jpg"},{"id":101420894,"identity":"b67246c3-a035-4783-b3c8-b5bd2b2aeed8","added_by":"auto","created_at":"2026-01-29 13:29:08","extension":"jpg","order_by":8,"title":"Figure 8","display":"","copyAsset":false,"role":"figure","size":158559,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cem\u003eFlow rates in L/m\u003c/em\u003e\u003csup\u003e\u003cem\u003e2\u003c/em\u003e\u003c/sup\u003e\u003cem\u003eh were measured over time in three scales of cyclical cake filtration with Culture 3 (Table 1). The filter areas tested were 0.001m\u003c/em\u003e\u003csup\u003e\u003cem\u003e2\u003c/em\u003e\u003c/sup\u003e\u003cem\u003e (Nutsche), 0.006m\u003c/em\u003e\u003csup\u003e\u003cem\u003e2\u003c/em\u003e\u003c/sup\u003e\u003cem\u003e (FUNDALOOP® 50), and 0.018m\u003c/em\u003e\u003csup\u003e\u003cem\u003e2\u003c/em\u003e\u003c/sup\u003e\u003cem\u003e (FUNDALOOP® 200). 80% Celpure C300 was added per biomass, as well as INDION 810 at a 1:5 ratio relative to DE. The error bars represent standard deviations.\u003c/em\u003e\u003c/p\u003e","description":"","filename":"Fig.8.jpg","url":"https://assets-eu.researchsquare.com/files/rs-8605471/v1/6920ab4a18439e5ea124e1bc.jpg"},{"id":101420888,"identity":"d9da9cfd-d42c-4900-9cd1-cfb751c23596","added_by":"auto","created_at":"2026-01-29 13:29:06","extension":"jpg","order_by":9,"title":"Figure 9","display":"","copyAsset":false,"role":"figure","size":111144,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cem\u003eTurbidity, product yield (based on IgG), DNA, and HCP levels were compared over three scales of cyclical cake filtration with Culture 3 (Table 1). Turbidity was measured in NTU and the remaining parameters were measured in a relative concentration to the feed material by the same methodology as described in Fig. 4. Celpure C300 and INDION 810 were added as in Fig. 8. The FUNDALOOP® 200 was only tested once for turbidity and yield; all other tests were conducted three times. The error bars represent standard deviations.\u003c/em\u003e\u003c/p\u003e","description":"","filename":"Fig.9.jpg","url":"https://assets-eu.researchsquare.com/files/rs-8605471/v1/5b18af3cf44d68c64d8626c4.jpg"},{"id":102397193,"identity":"bf183d3e-13c2-4c05-ae29-670199edc698","added_by":"auto","created_at":"2026-02-11 10:08:42","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":4281127,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-8605471/v1/70910276-0632-4f41-ae23-485f6b3edcf9.pdf"}],"financialInterests":"Competing interest reported. Financial interests: Author CJJ and TB are employees of DrM, which manufactures the filtration system evaluated in this study. \nNon-financial interests: The authors have no non- financial competing interests to declare.","formattedTitle":"Increasing CHO cell harvest efficiency with cyclical cake filtration","fulltext":[{"header":"Key Points","content":"\u003cul\u003e\n \u003cli\u003eCyclical cake filtration offers an alternative to depth and alluvial filtration\u003c/li\u003e\n \u003cli\u003eMore efficient cell harvest at reduced filter area, lower waste and lower cost\u003c/li\u003e\n \u003cli\u003eHigher product recovery was achieved with cyclical cake filtration\u003cbr\u003e\u0026nbsp;\u003c/li\u003e\n\u003c/ul\u003e"},{"header":"Introduction","content":"\u003cp\u003eWith advancements in upstream design and the growing demand for pharmaceuticals and therapies, biomanufacturers are prioritizing efficient, high yield production processes (Auerbach 2025; Sharma et al. 2017; Thakur et al. 2020). After culture is grown in a bioreactor, clarification functions as the first step of removing cell debris, particulates, and ideally some impurities from the upstream harvest. Sometimes referred to as \u0026ldquo;midstream,\u0026rdquo; clarification acts as the gateway into downstream processing, known to be one of the biggest cost-intensive steps of bioproduction (Sharma et al. 2017; Rooij et al. 2019). Clarification is often a multi-stage operation, where primary clarification removes larger cell debris and secondary clarification removes finer impurities and particulates.\u003c/p\u003e \u003cp\u003eFor primary clarification, centrifuges and depth filters are popular options. For secondary clarification, typically another depth filter is used (Sharma et al. 2017). Disposable depth filters, although used widespread in industry \u0026ndash; by 79% of respondents in the 16th Annual Report and Survey of Biopharmaceutical Manufacturing Capacity and Production \u0026ndash; have various shortcomings especially with large scale manufacturing. Upstream variability or high cell densities can lead to dramatic decrease of flow rates and early fouling of depth filters (Minow et al. 2014). Cells commonly are captured directly within the tortuous path of the thick depth filter media (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003ea), forming a gel-like layer on the top. Nonuniform flow and pressure increases may trigger cells to break free and cause turbidity breakthroughs, jeopardizing reliable clarity requirements for filtrate (Nejatishahidein et al. 2022). Furthermore, once a depth filter membrane is clogged, the cartridge is considered spent and must be replaced. To ensure a complete batch can be processed without risk of early fouling, various sources in the literature suggest purposefully oversizing the filter area, referencing safety factors between 1.2 and 2.3 (Laska et al. 2005; Thakur et al. 2020; Dryden et al. 2021). Compensating this way may solve the problem of unreliable results when scaling up, but it comes at the cost of efficiency.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eOne solution to improve the throughput of filters is to incorporate a filter aid, such as diatomaceous earth (DE). DE is made up of micron-scale diatom skeletons that have a unique incompressible and highly porous structure, and there are multiple commercially available types for GMP manufacturing that are basically inert (Daumke et al. 2019). Although adsorptive depth filters often include DE inside their cartridges, alluvial depth filters leverage DE by adding it directly to the feed stream (also known as body feed filtration) to form a filter cake on top of the media (Scott 2017; Meierrieks et al. 2023). Alluvial filters protect the filter media and dramatically reduce the risk of turbidity breakthrough, as the porous cake serves as the first line of defense to capture cell debris and particulates (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003eb). This additional filtration technique to remove process solids reduces the overall requirement of filter area, as the media is less prone to clogging. Some instances have shown a 3.5-fold increase in filter capacity when comparing alluvial filtration to centrifugation and depth filtration (Meierrieks et al. 2023). Filter capacity is directly correlated with economic and ecologic impacts, as less filter area is required, and thus, less waste generated. Despite these advantages, filter area remains a limiting factor to the scale-up of alluvial filtration, along with the volume of the cake (Ledergerber 2022; Daumke et al. 2019). Once a capsule is filled, it is considered waste and must be replaced with a new capsule.\u003c/p\u003e \u003cp\u003eA recent advancement aimed at improving efficiency of cake filtration is the development of a single-use multi-cycle cake filter. This technology utilizes the same advantages of a porous filter cake to increase filter throughput and lifespan, with the added benefit of cloth regeneration step over multiple cycles (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003ec). Once the cake reaches a maximum volume, a sterile backflush step is automatically performed, cleaning the filter cloth and restoring its original flow rate. This is run in a cyclical function and can be repeated up to 50 times, massively increasing the capacity of the filter. Furthermore, the filter media is significantly thinner than depth filter media, allowing for higher flow rates, significantly reduced risk of fouling, and no requirement for preflushing.\u003c/p\u003e \u003cp\u003eThe selection of what technology to use for clarification is not only a nuanced challenge but can also have vast economic impacts on a process. Throughput, scalability, and material waste must be considered to improve overall efficiency of a harvesting step. This study investigates cyclical cake filtration and establishes a comparison to depth and alluvial filtration with respect to performance metrics, scalability, and practical implementation.\u003c/p\u003e"},{"header":"Materials and Methods","content":"\u003cp\u003eAll experiments were conducted using an immunoglobulin G (IgG)-expressing ExpiCHO-S cell line (Gibco). The cultures that were used are summarized in Table\u0026nbsp;\u003cspan refid=\"Tab1\" class=\"InternalRef\"\u003e1\u003c/span\u003e. For inoculum production and the subsequent batch and fed-batch cultivations, Efficient-Pro Medium (Gibco) supplemented with 6 mmol L\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e L-glutamine and 0.1% Anti-Clumping Agent was used as basal medium. Efficient-Pro Feed 1 (Gibco) for cultivation 2 and Efficient-Pro Feed 2 (Gibco) for cultivation 3 were used throughout the fed-batch cultivations. Cells were cultivated in bioreactors with 5 to 10 L working volume.\u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab1\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 1\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003e\u003cem\u003eExpiCHO cell cultures used throughout the study. Culture 1 was used for\u003c/em\u003e Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003e, \u003cem\u003eculture 2 for\u003c/em\u003e Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003e, \u003cem\u003eand culture 3 for\u003c/em\u003e Figs.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003e\u0026ndash;\u003cspan refid=\"Fig9\" class=\"InternalRef\"\u003e9\u003c/span\u003e\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"6\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c5\" colnum=\"5\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c6\" colnum=\"6\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\"\u003e \u003cp\u003eCultivation #\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003eType\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c3\"\u003e \u003cp\u003eCell density [mio cells/ml]\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c4\"\u003e \u003cp\u003eViability\u003c/p\u003e \u003cp\u003e[%]\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c5\"\u003e \u003cp\u003eCrude Turbidity [NTU]\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c6\"\u003e \u003cp\u003eBiomass\u003c/p\u003e \u003cp\u003e[%]\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e1\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eBatch\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e9.5\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e99.2\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e454\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e1.0\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e2\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eFed-batch\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e8.5\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e75.0\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e950\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e3.0\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e3\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eFed-batch\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e10.7\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e96.3\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e870\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e2.8\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003cp\u003eThe first two cultures were used for preliminary tests that were conducted once, while the third culture was used for the final comparison and scale-up where test conditions were repeated three times (with few exceptions as mentioned). All cultures were placed in the fridge after completing the cultivation and filtered on the subsequent day(s).\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eFor the cyclical cake filtration, a proprietary polypropylene filter cloth was used. All small-scale experiments were performed in a Nutsche that is shown in Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003ea. A Nutsche consists of a cylindrical body with 200 ml volume underneath which the filter medium with 0.001 m\u003csup\u003e2\u003c/sup\u003e cross-sectional area is placed on a porous plate substrate. Pressurized air is applied from the top, and the pressure is set to a constant value via a regulator. The scale-up tests in Figs.\u0026nbsp;\u003cspan refid=\"Fig9\" class=\"InternalRef\"\u003e9\u003c/span\u003e and 10 were performed with the FUNDALOOP\u0026reg; skid shown in Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003ec and single-use bags having 0.006 m\u003csup\u003e2\u003c/sup\u003e and 0.018 m\u003csup\u003e2\u003c/sup\u003e cross sectional filter area (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003eb).\u003c/p\u003e \u003cp\u003eFor the alluvial filtration, cellulose and diatomaceous earth-based media with a retention range of 10\u0026ndash;30 \u0026micro;m and 0.1\u0026ndash;0.3 \u0026micro;m were used for the primary and secondary clarification step, respectively. The finer medium was charged with cationization agents. The tested filter area was 0.001 m\u003csup\u003e2\u003c/sup\u003e.\u003c/p\u003e \u003cp\u003eFor the depth filtration, cellulose and diatomaceous earth-based dual layer media with a retention range of 0.5\u0026ndash;5 \u0026micro;m and 0.05\u0026ndash;3 \u0026micro;m were used for the primary and secondary clarification step, respectively. The latter again had an anion exchange functionality to adsorb impurities. Lab filter capsules with a filter area of 0.0025 m\u003csup\u003e2\u003c/sup\u003e were used for the tests, with the exception of the test in Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003e, where 0.001 m\u003csup\u003e2\u003c/sup\u003e was used. A Watson Marlow peristaltic pump with pump head 114DV and 3.2 mm inner diameter tubing was used to establish a constant flow, and a PendoTech single-use pressure sensor was used to measure the pressure. The pressure reading was monitored using a Levitronix LCO-i100 process control unit. To calculate the depth filter sizing in Fig.\u0026nbsp;\u003cspan refid=\"Fig8\" class=\"InternalRef\"\u003e8\u003c/span\u003e, the P_max method was used (Merck KGaA 2020).\u003c/p\u003e \u003cp\u003eThe flow rates in all filtration methods were measured indirectly via a scale and taking periodic weight measurements, assuming that the density of the product is identical to the one of water.\u003c/p\u003e \u003cp\u003eFor the alluvial and cyclical cake filtration, diatomaceous earth (DE) was added to the feed prior to the filtration. For alluvial filtration, 65% per biomass Celpure C300 was added, which is within the typical range of 50\u0026ndash;100% (van der Meer et al. 2014). This high-purity DE is specifically designed for biopharmaceutical applications and has a nominal permeability of 300 mDarcy. For the cyclical cake filtration, 65\u0026ndash;100% per biomass was used. For Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003e, 65% was used to have a meaningful comparison with alluvial filtration. Celpure C300 was again used for all experiments with the exception of one process condition tested in Figs.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003e\u0026ndash;\u003cspan refid=\"Fig7\" class=\"InternalRef\"\u003e7\u003c/span\u003e where 80% per biomass Celpure S25E was used, having a nominal permeability of 25 mDarcy.\u003c/p\u003e \u003cp\u003eThe biomass was calculated from the total cell count and the average cell diameter, assuming a spherical shape of the cells. For cake washing, phosphate buffered saline (PBS) was used that was dissolved in water at a concentration of 9.55 g/L.\u003c/p\u003e \u003cp\u003eThe second trains of depth and alluvial filters (which are essentially normal depth filters due to the absence of DE) are typically charged in order to adsorb unwanted impurities such as deoxyribonucleic acid (DNA) and host cell proteins (HCP). In order to study the possibility of removing DNA and HCP, several ion exchange resins were compared in this study: INDION 810, a macroporous strongly basic type I anion exchange resin, INDION 820, a macroporous strongly basic type II anion exchange resin, INDION 830, a strong base type I anion exchange resin having quaternary ammonium function groups, and INDION 254, a strongly acidic cation exchange resin. Those ion exchange resins were contrasted with Daicel MabXPure that combines anion exchange (AEX) and size exclusion chromatography (SEC) and that has been featured in previous studies (Daumke et al. 2019, Daicel Bioseparations n.d.).\u003c/p\u003e \u003cp\u003eThe turbidity of the filtrate was measured using a WTW Turb 430 IR/T turbidity meter, and the cake thickness in cyclical cake filtration and alluvial filtration was measured using a caliper.\u003c/p\u003e \u003cp\u003eTo evaluate the filter performance, Immunoglobulin G (IgG), HCP, and DNA concentration were measured before and after filtration. Prior to the analysis the samples were centrifuged at 3000 g for 5 min to remove any remaining cells. IgG concentration was measured with the IgG Bio test kit (Roche Diagnostics) using the Cedex Bio Analyzer (Roche Diagnostics). For the DNA and HCP measurements, the samples were diluted according to the manufacturer\u0026rsquo;s recommendation. As the actual concentrations were unknown, three different dilutions were analyzed per sample. HCP concentration was measured using the CHO Host Cell Proteins ELISA Kit (RayBiotech). DNA concentration was measured using the Quant-iT Picogreen dsDNA Assay Kit (Invitrogen). Both HCP and DNA kits were evaluated using the Varioskan LUX spectrometer (Thermoscientific).\u003c/p\u003e"},{"header":"Results","content":"\u003cp\u003eIn a Nutsche filter, 125 ml sample sizes of Culture 2 (Table\u0026nbsp;\u003cspan refid=\"Tab1\" class=\"InternalRef\"\u003e1\u003c/span\u003e) were filtered using a depth filter, alluvial filter, and cyclical cake filter at a constant pressure of 1.5 bar, which was chosen to limit cell damage (Minow et al. 2014). Depth filters are not normally run in a pressure-control mode (Sharma 2017), however this was done for the sake of understanding the filtration mechanisms. Complete blockage of the depth filter occurred after filtration of 125 ml. Figure\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003e shows the flow rates in L/m\u003csup\u003e2\u003c/sup\u003eh (LMH) over time. As alluvial filtration and cyclical cake filtration exhibited a faster flow rate than depth filtration, data was extrapolated and displayed in respective dotted lines to show how the flow rate was expected to drop to an asymptotic plateau over time.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eFigure\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003e compares the effect of resins on product yield and the removal of impurities. Yield was calculated by measuring the concentrations and volumes before and after the filtration. The volume before the filtration was corrected by the volume occupied by cells. In Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003ea, five different ion-exchange resins were added to 160 ml samples of Culture 1 (Table\u0026nbsp;\u003cspan refid=\"Tab1\" class=\"InternalRef\"\u003e1\u003c/span\u003e) prior to filtration in the Nutsche with the cyclical cake filter cloth. Out of those tested, INDION 810 and MabXPure removed the most DNA and HCP and therefore were selected for further investigation in Figs.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003eb and \u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003ec.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eFigure\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003e \u003cem\u003eDNA, HCP, and IgG concentrations were compared with addition of various ion exchange resins. 100% refers to the initial concentration of DNA, HCP, and IgG in the feed. For all tests, 160 ml of Culture 1 and 100% Celpure C300 per biomass were filtered in the Nutsche with media from a cyclical cake filter. a) Five different ion-exchange resins were added to the feed. All INDION resins were added at ratios 1:2 relative to the weight of DE, and MabXPure was added at a ratio 1:5 to the feed. b) MabXPure was added in four volumetric ratios (1:50, 1:20, 1:10, 1:5) to the feed. c) INDION 810 was added in two ratios (1:5 and 1:2) to the weight of DE added, each for two mixing times (5 minutes or 30 minutes).\u003c/em\u003e\u003c/p\u003e \u003cp\u003eAll tests henceforth were done using Culture 3 (Table\u0026nbsp;\u003cspan refid=\"Tab1\" class=\"InternalRef\"\u003e1\u003c/span\u003e). For depth filtration, a capacity of 100 ml was determined via \u003cem\u003eP_max\u003c/em\u003e method, and this filtration amount was used throughout the tests in Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003e\u0026ndash;\u003cspan refid=\"Fig7\" class=\"InternalRef\"\u003e7\u003c/span\u003e. For alluvial and cyclical cake filtration, 125 ml were filtered based on maximum cake thickness considerations. An exception was made for the condition using Celpure S25E and a finer filter cloth, where only 75 ml of culture was filtered.\u003c/p\u003e \u003cp\u003eDepth and alluvial filtrations were divided into separate stages for primary and secondary clarification, in which filtrate from the first clarification was collected and measured before being fed into the second filter for further clarification. Both depth filters and alluvial filters were flushed with the prescribed washing liquid prior to filtration (135ml and 50ml, respectively). Alluvial filtration was performed using a peristaltic pump for the first test and then was switched to filtration in a Nutsche for the second and third test due to issues with DE sedimentation in the peristaltic pump tubing. Cyclical cake filters were not pre-flushed as per the operating instructions and were tested using four different process parameters in only one stage.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eFigure\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003e \u003cem\u003eFiltrate NTU was measured after processing Culture 3 with cyclical cake filtration in a single stage for four process conditions (always using 80% per biomass), and in two stages with alluvial filtration and depth filtration. Depth filters were run with a peristaltic pump at 10ml/min, or 240 LMH, and alluvial filters were run with a peristaltic pump at 15ml/min, or 900 LMH, as well as a Nutsche filter. Cyclical depth filters were processed in the Nutsche with a constant pressure of 1.5 bar. The following additions were tested: Celpure C300, Celpure S25E with a finer cloth, Celpure 300 with INDION 810 (1:5 ratio relative to DE), and Celpure C300 with MabXPure (1:10 ratio relative to feed). The error bars represent standard deviations.\u003c/em\u003e\u003c/p\u003e \u003cp\u003eIn Fig.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e6\u003c/span\u003ea, the same four testing conditions for cyclical cake filtration were compared to alluvial filtration and depth filtration in regard to product yield, both before and after washing with PBS. All tested scenarios showed a significant amount of product can be recovered with a washing step. Product yield was lowest after depth filtration and highest after cyclical cake filtration. The condition which resulted the highest product yield for cyclical cake filtration was with Celpure C300 alone. Figure\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e6\u003c/span\u003eb compares the impurity levels of the filtrate. Depth filters showed the lowest level of both DNA and HCP. Alluvial filtration and cyclical depth filtration with the addition of MabXPure exhibited similar impurity profiles. The DNA level after filtration with Celpure C300 and S25E were higher than 100% because the filtration induces some amount of cell damage that gives rise to an increase in the DNA concentration.\u003c/p\u003e \u003cp\u003eFigure 6 \u003cem\u003eProduct yield (based on IgG), DNA, and HCP levels from Culture 3 (Table 1) were measured using the same testing conditions as in Fig. 5. a) IgG levels were measured in the filtrate before and after a PBS wash. For cyclical cake filtration and alluvial filtration, 20% PBS wash per feed volume was used. For depth filtration, 25% PBS wash per feed volume was used. b) DNA and HCP levels were measured in the filtrate before washing. The error bars represent standard deviations. \u003c/em\u003e \u003c/p\u003e \u003cp\u003eFigure\u0026nbsp;\u003cspan refid=\"Fig7\" class=\"InternalRef\"\u003e7\u003c/span\u003e estimates the required filtration area to process 2\u0026rsquo;000 L of cell harvest for each filter type. Depth filtration required the most filter area, as was calculated using a \u003cem\u003eP_max\u003c/em\u003e calculation, assuming a maximum duration of five hours processing time and a maximum pressure of 1.5 bar (Thakur et al. 2020; Cytiva 2023). Data shown for alluvial filtration reflects a maximum cake thickness of 18mm (Filtrox Group n.d.), and was extrapolated from earlier Nutsche trials, as cake thickness scales linearly with the filtered amount. Less filter area is required for alluvial filtration compared to depth filtration. For cyclical cake filtration, it was determined from previous Nutsche tests that 100 ml per filtration with a 0.001m\u003csup\u003e\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e\u003c/sup\u003e filter area yielded a cake of approximately 9.5 mm, allowing for a successful discharge. It is assumed that the first quarter of the filtrate is recycled back as prefiltrate, reducing the amount that can be filtered with 0.001m\u003csup\u003e\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e\u003c/sup\u003e to 75 ml. Using these values to scale up to two commercial-sized FUNDALOOP\u0026reg; filters with a combined filter area of 0.88m\u003csup\u003e\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e\u003c/sup\u003e, it is calculated that 66 L can be filtered per cycle, allowing for the filtration of 2\u0026rsquo;000 L in 30 cycles.\u003c/p\u003e \u003cp\u003e Figure 7 \u003cem\u003eThe required filter area to process 2,000 L of CHO culture was calculated for the three filter methods. It was assumed that secondary clarification for alluvial and depth filtration requires half of the filter area as primary clarification (Zustiak et al. 2022; Merck KGaA 2017). A typical safety factor of 50% was included for alluvial and depth filtration (Dryden et al. 2021). \u003c/em\u003e\u003c/p\u003e \u003cp\u003eFigure\u0026nbsp;\u003cspan refid=\"Fig8\" class=\"InternalRef\"\u003e8\u003c/span\u003e compares flow rates between three scales for cyclical cake filtration: the Nutsche, FUNDALOOP\u0026reg; 50, and FUNDALOOP\u0026reg; 200. The tested volumes at the three scales were 125 ml, 750 ml, and 2,250 ml, respectively. With the exception of a brief initial surge in flow recorded in the Nutsche, the flow rates exhibited similar patterns and values between scales.\u003c/p\u003e \u003cp\u003e Figure 8 \u003cem\u003eFlow rates in L/m2h were measured over time in three scales of cyclical cake filtration with Culture 3 (Table 1). The filter areas tested were 0.001m2 (Nutsche), 0.006m2 (FUNDALOOP® 50), and 0.018m2 (FUNDALOOP® 200). 80% Celpure C300 was added per biomass, as well as INDION 810 at a 1:5 ratio relative to DE. The error bars represent standard deviations. \u003c/em\u003e\u003c/p\u003e \u003cp\u003eFiltrate quality parameters were compared between the same three scales for cyclical cake filtration in Fig.\u0026nbsp;\u003cspan refid=\"Fig9\" class=\"InternalRef\"\u003e9\u003c/span\u003e. Product yield was consistent and similar across all scales. DNA and HCP were less consistent than yield, but within range of each other, especially considering the larger standard deviations associated with the measurement. Turbidity was lower in the Nutsche than the FUNDALOOP\u0026reg; 50 or 200.\u003c/p\u003e \u003cp\u003e Figure 9 \u003cem\u003eTurbidity, product yield (based on IgG), DNA, and HCP levels were compared over three scales of cyclical cake filtration with Culture 3 (Table 1). Turbidity was measured in NTU and the remaining parameters were measured in a relative concentration to the feed material by the same methodology as described in Fig. 4. Celpure C300 and INDION 810 were added as in Fig. 8. The FUNDALOOP® 200 was only tested once for turbidity and yield; all other tests were conducted three times. The error bars represent standard deviations.\u003c/em\u003e\u003c/p\u003e"},{"header":"Discussion","content":"\u003cp\u003eA critical element of efficiency in bioproduction involves how quickly a batch can be processed. Faster flow rates during cell harvesting and clarification is one way to decrease processing time and increase efficiency. As shown in Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003e, flow rates with depth filter media slowed down quickly and approached an asymptotic plateau faster than alluvial filtration or cyclical cake filtration. This can be explained by the mechanism of capture; cells and debris are retained by the thick depth filter media itself, which leads to rapid clogging. After filtering 125mls of Culture 1, the media was almost completely fouled.\u003c/p\u003e \u003cp\u003eIn contrast, no clogging occurred in alluvial filtration, as the cells and debris are chiefly held back by the filter cake, which protects the filter media and maintains a higher flow rate. Had the filtration been continued, the flow rate would have also continued to drop due to the increased resistance of the cake, but even in the asymptotic plateau the flow rates would have been considerable. Unlike in depth filtration, the capacity is defined either by the maximum cake thickness \u0026ndash; a physical limitation of the filter \u0026ndash; or a specified minimum flow rate.\u003c/p\u003e \u003cp\u003eThe mechanism of capture in cyclical cake filtration is similar to alluvial filtration, but the flow rates are even higher due to the thin filter media having a lower resistance than the thick and tortuous depth filter media. The maximum capacity per cycle could be even higher than in alluvial filtration, however, the capacity is usually limited by the maximum allowable cake thickness \u0026ndash; typically between 10-12mm. Nonetheless, the overall capacity is extensively increased with the allowance of 50 cycles to be performed using one cloth.\u003c/p\u003e \u003cp\u003eOverall, Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003e demonstrates that cyclical cake filtration significantly decreases the process risk during the primary clarification step. The incorporation of filter aid eliminates the risk of clogging \u0026ndash; one of the key downfalls of depth filters that often leads to a partial loss of product titers. Additionally, there is no risk of overfilling filters, as is a concern in alluvial systems, because a more challenging filtration is tackled with an increased number of filtration cycles. Variabilities in the production or feed streams can therefore be effectively handled without the need of adding more filter area.\u003c/p\u003e \u003cp\u003eWhile this study shows that depth filters struggled to achieve high throughput and would require an exorbitant amount of filter area at scale, they are known to capture a large portion of impurities with isoelectric bonds by charging their membrane. However, this advantage could be rivaled in alternative filtration methods by the addition of ion exchange resins to the feed. Figure\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003eb and \u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003ec show two resin options for reduction of impurities in cyclical cake filtration: MabXPure and INDION 810. The addition of MabXpure at a 1:5 ratio of resin to feed reduced the DNA and HCP levels the most, however this did come at an expense to the product yield. Reducing the ratio of MabXPure to 1:10 increased the product yield to approximately 90% while also removing a majority of the impurities (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003eb). Furthermore, INDION 810 resulted in a higher product yield of approximately 95%, and still removed the majority of DNA and HCP even with a low ratio, as long as the resin was incubated for 30 minutes prior to filtration. IgG levels were unaffected by INDION 810 ratio or incubation time.\u003c/p\u003e \u003cp\u003eWhile a reduced impurity profile helps to reduce the burden of further downstream steps, other characteristics of high filtrate quality include filtrate clarity and product yield. These parameters were compared between cyclical cake filtration, alluvial filtration, and depth filtration in Figs.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003e and \u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e6\u003c/span\u003e. It is important to note that alluvial filtration and depth filtration were each run in a two-stage filter train, standardly referred to as primary and secondary clarification. The data shown from cyclical cake filtration is obtained from only one round of filtration.\u003c/p\u003e \u003cp\u003eFor a standard filtration where 80% Celpure C300 per biomass was used, the turbidity of cyclical cake filtration roughly matched the turbidity achieved after the first clarification with alluvial and depth filters. With the addition of MabXPure in a 1:10 ratio of resin to feed and 80% Celpure C300 per biomass, the turbidity could be significantly reduced to less than 5 NTU and rivaled the filtrate clarity of depth filtration after primary and secondary clarification (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003e). The possibility to achieve similar clarity levels in one step versus two could have extreme impacts on cost, time, and complexity of cell harvesting in a bioprocess. Another cyclical cake condition shown in Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003e that reduced filtrate NTU was using a finer DE grade (Celpure S25E) and a finer cloth. Compared to Celpure C300, the smaller size and lower permeability of Celpure S25E aids in capturing smaller micron sized particles. This DE substitution yielded a filtrate NTU of approximately 12 in a single step, similar to the level which was reached after the second stage of alluvial clarification. This trade-off reflects a well-documented phenomenon in filtration systems, where tighter filter media typically yield improved filtrate clarity at the expense of increased pressure buildup and reduced flow rate. These parameters are interdependent and therefore must be carefully balanced when optimizing filtration performance.\u003c/p\u003e \u003cp\u003eIn addition to reducing the filtrate NTU, MabXPure also significantly improved the impurity profile of the filtrate (Fig.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e6\u003c/span\u003eb). The levels of DNA and HCP in the filtrate from cyclical cake filtration with the addition of MabXPure were nearly identical to that of alluvial filtration. Although depth filtration still resulted in the lowest impurity profile, the addition of MabXPure provides an alternative option to reduce DNA and HCP without depth filtration. It is prudent to note that data shown in Fig.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e6\u003c/span\u003eb was measured pre-washing with PBS, and future studies must be done to measure the impurity levels post washing to understand if this would increase DNA or HCP levels. It is likely that the addition of ion exchange resins in the cake would tightly bond impurities and therefore retain them well even during washing, but this needs to be investigated further.\u003c/p\u003e \u003cp\u003eIt also needs to be noted that due to the aforementioned challenges with DE sedimentation in the peristaltic pump tubing, the second and third alluvial filtrations were carried out in a Nutsche. These tests were therefore subjected to higher flow rates than are suggested by the manufacturer due to operational limitations of the Nutsche. It is probable that if alluvial filtration had been conducted with slower flow rates, longer contact time with the filter media would have resulted in more effective capture of DNA and HCP, similar to the levels shown with depth filtration. This would be best confirmed in a future study.\u003c/p\u003e \u003cp\u003eAlthough the surface charge on depth filter media can aid in impurity removal, it is non-specific and can therefore capture any molecule that it has an electrostatic interaction with. This can include the product of interest, in turn resulting in a lower product yield. This is exemplified in Fig.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e6\u003c/span\u003ea, with depth filters having the lowest product yield compared to alluvial and cake filtration. Product can also be retained by the filter cake and cloth in cyclical cake filtration and alluvial filtration, but the effect of this is less pronounced. The highest product yield was observed after cyclical cake filtration with the addition of Celpure C300 without ion exchange resin. The finer DE grade and additional resins resulted in a slightly lower product yield, albeit higher than depth filtration and alluvial filtration.\u003c/p\u003e \u003cp\u003eWhen considering all results of Figs.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003e and \u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e6\u003c/span\u003e, cyclical cake filtration with the addition of Celpure C300 and MabXPure was able to achieve very similar results as two-stage depth or alluvial filters in terms of impurity removal and filtrate clarity, albeit at a higher product yield.\u003c/p\u003e \u003cp\u003eThese results were obtained using IgG as the model product. Because different molecules exhibit distinct isoelectric points (pI) and adsorptive properties, product yield may vary substantially depending on the process conditions. Therefore, thorough optimization is necessary to determine which filter aids and additives, if any, are most suitable for each product and process.\u003c/p\u003e \u003cp\u003eFurthermore, although these findings demonstrate that the addition of MabXPure can reduce impurity levels in the filtrate while achieving a high product yield, the cost of resin additives could offset the benefits. Harvest conditions and requirements should be evaluated individually to determine if resin addition is a valuable and economic option.\u003c/p\u003e \u003cp\u003eIt can be argued that filtrate quality is the most critical measure of filtration performance, but practical implementation will often depend on a proper scale-up evaluation. In Fig.\u0026nbsp;\u003cspan refid=\"Fig7\" class=\"InternalRef\"\u003e7\u003c/span\u003e, data was extrapolated from small scale trials to estimate required filtration area for a 2,000 L batch, commonly used as a commercial scale production volume when using single-use bioprocessing equipment. Alluvial filtration significantly reduces filter area requirement compared to depth filters, as the filter cake extends the life of the media. Further supporting this, Minow et al. (2014) reported that scaling up depth filters to commercial levels often times does not make financial sense due to low filtration capacities per filter area. However, both depth filters and alluvial filters are typically oversized and installed with a safety factor. In alluvial filters, this safety factor is intended to avoid filter filling, whereas in depth filters it primarily prevents filter blockage.\u003c/p\u003e \u003cp\u003eCyclical cake filtration leverages the benefit of a protective cake layer in alluvial filtration even further with its backflush step and regenerative nature, causing the required surface area to dramatically drop. Moreover, there is no need for an additional safety factor, as each filter can withstand up to 50 cycles. The 30 cycles calculated to process 2,000 L in Fig.\u0026nbsp;\u003cspan refid=\"Fig7\" class=\"InternalRef\"\u003e7\u003c/span\u003e would result in 66% safety factor over the maximum allowable number of cycles. Overall, cyclical cake filtration allows for 98% and 96.5% filter area savings compared to depth and alluvial filtration, respectively.\u003c/p\u003e \u003cp\u003eRequired filtration area directly correlates to operational cost, potential for extractables and leachables to compromise a batch, and quantity of plastics and single-use materials that are discarded after each batch. Furthermore, capital equipment requirement and therefore cleanroom space also increase with higher consumable usage. Implementation of cyclical cake filtration as a harvest method can have substantial economic and environmental benefits, creating a leaner and more efficient bioprocess step.\u003c/p\u003e \u003cp\u003eIt needs to be mentioned that this study did not conduct any optimization on the depth and alluvial filtrations and simply compared models that are standard in the industry for both. In recent years, those competing technologies have been continuously developed further by manufacturers, resulting in products on the market that achieve better results in terms of capacity, yield, and removal of impurities. However, those filters still have the same drawbacks (can only be used once until clogging, are not regenerable, and require a large filter area), and it is expected that general trends in the data remain the same.\u003c/p\u003e \u003cp\u003eScalability between individual cyclical cake filtration systems is shown in Figs.\u0026nbsp;\u003cspan refid=\"Fig8\" class=\"InternalRef\"\u003e8\u003c/span\u003e and \u003cspan refid=\"Fig9\" class=\"InternalRef\"\u003e9\u003c/span\u003e, detailing flow rates and filtrate quality. Between the Nutsche system and FUNDALOOP\u0026reg; 200, there is an 18-fold increase in the filter area. Flow rates across all three systems behave similarly, apart from an initial high flow rate in the Nutsche as the operational mechanism differs and 1.5 bar is achieved instantly. In the FUNDALOOP\u0026reg; systems, pressure builds up slower to account for bag filling and formation of the cake, resulting in a lower initial flow rate. After pressure is established, flow patterns between the three systems are very similar.\u003c/p\u003e \u003cp\u003eThe quality profile of filtrate was comparable between the three scales as well. There were slight variations in impurity levels, but product yield was relatively unaffected by the difference in scales. The turbidity was lower in the Nutsche system, however, which can be explained by the absence of turbulence inside the filter vessel. In the FUNDALOOP\u0026reg; system, upwards flow of feed creates an added element of homogenization and filtration is not working in the direction of gravity. The observed variation of filtrate clarity is likely a function of this mechanistic difference as opposed to a change to the filtration area. Within the different FUNDALOOP\u0026reg; sizes, the turbidity values are stable and consistent.\u003c/p\u003e \u003cp\u003eA larger system, the FUNDALOOP\u0026reg; 3000, is also commercially available and exhibits approximately a 24-fold increase compared to the FUNDALOOP\u0026reg; 200. While tests with the FUNDALOOP\u0026reg; 3000 were outside the scope of this study, the presented data likely supports the scalability of this system as well.\u003c/p\u003e"},{"header":"Conclusion","content":"\u003cp\u003eThis study demonstrates that cyclical cake filtration delivers a highly efficient and scalable alternative to depth and alluvial filtration for conventional CHO harvest clarification, providing faster flow rates, higher product yield, reduced fouling, and significantly lower filtration area requirements. The use of filter aids protects the filter medium, eliminating the danger of clogging that is a key disadvantage of depth filters. The thin filter medium and regenerative cycling improve throughput, while the incorporation of ion exchange resins such as MabXPure or INDION 810 enables substantial removal of DNA and HCP \u0026ndash; achieving filtrate clarity and impurity levels comparable to multi-stage depth or alluvial filtration in a single unit operation. Additionally, performance can be tailored by adjusting DE grade, cake thickness, pressure limit, resin type, resin ratio, and incubation time, offering process flexibility for optimization. Collectively, these results highlight cyclical cake filtration as a robust and cost-effective technology to enhance midstream bioprocess efficiency.\u003c/p\u003e \u003cp\u003e \u003cb\u003eAuthor Contribution Statement\u003c/b\u003e \u003c/p\u003e \u003cp\u003eCJJ performed part of the filtration tests, helped with data analysis and interpretation, wrote the majority of the manuscript, and compiled references. TB designed the study, executed the filtration tests, performed the data analysis, and helped writing the manuscript. AM and AG did the cultivations and executed the IgG, DNA, HCP measurements and analyses. JO coordinated the efforts from the ZHAW and helped with the data analyses and interpretation. All authors read and approved the manuscript.\u003c/p\u003e"},{"header":"Declarations","content":"\u003cp\u003e\u003ch2\u003eCompeting Interests\u003c/h2\u003e\u003cp\u003eFinancial interests: Author CJJ and TB are employees of DrM, which manufactures the filtration system evaluated in this study. Non-financial interests: The authors have no non- financial competing interests to declare.\u003c/p\u003e\u003c/p\u003e\u003ch2\u003eFunding\u003c/h2\u003e \u003cp\u003eNo funding was received to assist with the preparation of this manuscript.\u003c/p\u003e\u003ch2\u003eAuthor Contribution\u003c/h2\u003e\u003cp\u003eCJJ performed part of the filtration tests, helped with data analysis and interpretation, wrote the majority of the manuscript, and compiled references. TB designed the study, executed the filtration tests, performed the data analysis, and helped writing the manuscript. AM and AG did the cultivations and executed the IgG, DNA, HCP measurements and analyses. JO coordinated the efforts from the ZHAW and helped with the data analyses and interpretation. All authors read and approved the manuscript.\u003c/p\u003e\u003ch2\u003eAcknowledgement\u003c/h2\u003e\u003cp\u003eWe would like to thank Samuel Schneider for his support during the HCP and DNA analysis, as well as Dardan Qereti for the support during experiments.\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\u003cli\u003e\u003cspan\u003eth Annual Report and Survey of Biopharmaceutical Manufacturing Capacity and Production (2019) BioPlan Associates, Inc. www.bioplanassociates.com/16th Accessed 12 Nov 2025\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eAuerbach M (2025) The bioprocess revolution: how technology and trends are reshaping pharmaceutical manufacturing. American Pharm Review. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://www.americanpharmaceuticalreview.com/Featured-Articles/618316-The-Bioprocess-Revolution-How-Technology-and-Trends-are-Reshaping-Pharmaceutical-Manufacturing\u003c/span\u003e\u003cspan address=\"https://www.americanpharmaceuticalreview.com/Featured-Articles/618316-The-Bioprocess-Revolution-How-Technology-and-Trends-are-Reshaping-Pharmaceutical-Manufacturing\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e/ Accessed 22 Oct 2025\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eCytiva (2023) Stax\u0026trade; mAx Clarification Platform. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://cdn.cytivalifesciences.com/api/public/content/kfGqaqgYvEi8m_T7wAWa1A-pdf\u003c/span\u003e\u003cspan address=\"https://cdn.cytivalifesciences.com/api/public/content/kfGqaqgYvEi8m_T7wAWa1A-pdf\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e Accessed 12 Nov 2025\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eDaicel Bioseparations (n d.) Static performance of MabXPure: application note #1. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://www.daicelchiral.com/wp-content/themes/daicel/assets/add_img/pdf/DAICEL-Bioseparations_Static-Performance-of-MabXpure-Application-Note-1.pdf\u003c/span\u003e\u003cspan address=\"https://www.daicelchiral.com/wp-content/themes/daicel/assets/add_img/pdf/DAICEL-Bioseparations_Static-Performance-of-MabXpure-Application-Note-1.pdf\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e Accessed 17 Dec 2025\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eDaumke R, Medvedev V, Albano T, Rousset F (2019) Alluvial filtration: an effective and economical solution for midstream application. In: Eibl R, Eibl D (eds) Single-Use Technology in biopharmaceutical manufacture, 2nd edn. Wiley, Inc, pp 271\u0026ndash;277\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eDryden W, Larsen L, Britt D, Smith M (2021) Technical and economic considerations of cell culture harvest and clarification technologies. Biochem Eng J 167:107892. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003edoi.org/10.1016/j.bej.2020.107892\u003c/span\u003e\u003cspan address=\"10.1016/j.bej.2020.107892\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eFiltrox Group (n d.) PURADISC\u0026reg; SU technical data sheet. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://www.filtrox.com/applications/filtration-in-life-science/single-use/\u003c/span\u003e\u003cspan address=\"https://www.filtrox.com/applications/filtration-in-life-science/single-use/\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e Accessed 16 Dec 2025\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLaska M, Brooks R, Gayton M, Pujar N (2005) Robust scale-up of dead end filtration: impact of filter fouling mechanisms and flow distribution. Biotechnol Bioeng 92(3):308\u0026ndash;320. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1002/bit.20587\u003c/span\u003e\u003cspan address=\"10.1002/bit.20587\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLedergerber B (2022) Upscaling: scaling of a filtration process. Filtrox AG. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://www.filtrox.com/wp-content/uploads/2022/08/AN_LS_005_Upscale_Jul2022_EN_note.pdf\u003c/span\u003e\u003cspan address=\"https://www.filtrox.com/wp-content/uploads/2022/08/AN_LS_005_Upscale_Jul2022_EN_note.pdf\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e Accessed 12 Oct 2025\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMeierrieks F, Pickl A, Wolff M (2023) A robust and efficient alluvial filtration method for the clarification of adeno-associated viruses from crude cell lysates. J Biotechnol 367:31\u0026ndash;41. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003edoi.org/10.1016/j.jbiotec.2023.03.010\u003c/span\u003e\u003cspan address=\"10.1016/j.jbiotec.2023.03.010\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMerck KGA (2017) Clarification of mammalian cell cultures by depth filtration. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://www.sigmaaldrich.com/deepweb/assets/sigmaaldrich/product/documents/350/422/pf1804en-mk.pdf\u003c/span\u003e\u003cspan address=\"https://www.sigmaaldrich.com/deepweb/assets/sigmaaldrich/product/documents/350/422/pf1804en-mk.pdf\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e Accessed 16 Dec 2025\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMerck KGA (2020) Pmax\u0026trade;/Tmax\u0026trade; constant flow rate test for depth filter sizing. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://www.sigmaaldrich.com/deepweb/assets/sigmaaldrich/product/documents/538/353/pmax-tmax-protocol-br4506en-mk.pdf\u003c/span\u003e\u003cspan address=\"https://www.sigmaaldrich.com/deepweb/assets/sigmaaldrich/product/documents/538/353/pmax-tmax-protocol-br4506en-mk.pdf\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e, accessed 16 Dec 2025\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMinow B, Egner F, Jonas F, Lagrange B (2014) High\u0026ndash;cell-density clarification\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eby single-use diatomaceous earth filtration (2025) Bioprocess Int. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://www.bioprocessintl.com/single-use/high-cell-density-clarification-by-single-use-diatomaceous-earth-filtration\u003c/span\u003e\u003cspan address=\"https://www.bioprocessintl.com/single-use/high-cell-density-clarification-by-single-use-diatomaceous-earth-filtration\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e Accessed 12 Oct\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eNejatishahidein N, Kim M, Jung S, Borujeni E, Fernandez-Cerezo L, Roush D, Borhan A, Zydney A (2022) Scale‐up issues for commercial depth filters in bioprocessing. Biotechnol Bioeng 119:1105\u0026ndash;1114. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1002/bit.28035\u003c/span\u003e\u003cspan address=\"10.1002/bit.28035\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eRooij J, DeConto J, Schaenzler G, Bauer D, Barre K, Duskin M, Kohli A, Watanabe K (2019) Upstream and downstream solutions for AAV manufacturing. Cell Gene Therapy Insights 5(Suppl 5):1017\u0026ndash;1029. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.18609/cgti.2019.110\u003c/span\u003e\u003cspan address=\"10.18609/cgti.2019.110\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eScott C (2017) Downstream disposables: the latest single-use solutions for downstream processing. BioProcess Int 15(1)i:1\u0026ndash;5\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSharma M, Raikar S, Srivastava S, Gupta S (2017) Examining single-use harvest clarification options: a case study comparing depth-filter turbidities and recoveries. BioProcess Int 15(2):40\u0026ndash;47\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eThakur G, Hebbi V, Parida S, Rathore A (2020) Automation of dead end filtration: an enabler for continuous processing of biotherapeutics. Front Bioeng Biotechnol 8:758. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.3389/fbioe.2020.00758\u003c/span\u003e\u003cspan address=\"10.3389/fbioe.2020.00758\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003evan der Meer T, Minow B, Lagrange B, Krumbein F, Rolin F (2014) Diatomaceous\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBioprocess Int 2(8):S25-S28 earth filtration: innovative single-use concepts for clarification of high-density mammalian cell cultures, Zustiak M, Smith M, Luo Q, Kruger J, Dryden W, Rivera M (2014) (2022) Simple, scalable single-stage harvest solution using chromatographic clarification technologies. Thermo Fisher Scientific Inc. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://documents.thermofisher.com/TFS-Assets%2FBPD%2FReference-Materials%2Fwhite-paper-single-use-harvest-solutions.pdf\u003c/span\u003e\u003cspan address=\"https://documents.thermofisher.com/TFS-Assets%2FBPD%2FReference-Materials%2Fwhite-paper-single-use-harvest-solutions.pdf\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e Accessed 16 Dec 2025\u003c/span\u003e\u003c/li\u003e\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":false,"hideJournal":true,"highlight":"","institution":"","isAcceptedByJournal":false,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true},"keywords":"CHO Cell Harvest, Clarification, Filtration, Impurity Removal, Scale-up, Processing efficiency","lastPublishedDoi":"10.21203/rs.3.rs-8605471/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-8605471/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003eEfficient clarification of bioreactor harvests remains a critical bottleneck in biomanufacturing, directly impacting throughput, product yield, and operational costs. Current industry practice typically relies on depth filters because they can effectively remove impurities, but they are also prone to fouling, require oversized filter areas, and can reduce product yield due to non-specific adsorption. As a response to some of those shortcomings, alluvial filtration was introduced which uses a diatomaceous earth (DE) cake to protect filter media, maintaining higher flow rates and extending filter life. Alluvial filters can reduce the required filter area, however, the true benefit of cake filtration can only be leveraged when using cyclical cake filtration. This technique uses multi-cycle cloth regeneration, achieving superior throughput and yield while maintaining filtrate clarity. Addition of ion exchange resins further allows for impurity reduction comparable to depth filtration without compromising product recovery. This study examines cyclical cake filtration and compares it with traditional and alluvial depth filtration regarding performance and scalability. Results indicate that cyclical cake filtration reduces required filter area and laboratory footprint, offering substantial economic and environmental benefits for commercial-scale production.\u003c/p\u003e","manuscriptTitle":"Increasing CHO cell harvest efficiency with cyclical cake filtration","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2026-01-29 13:29:01","doi":"10.21203/rs.3.rs-8605471/v1","editorialEvents":[{"type":"communityComments","content":0}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"456b44b0-4f73-423d-8842-3425ad2c9aa4","owner":[],"postedDate":"January 29th, 2026","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"posted","subjectAreas":[],"tags":[],"updatedAt":"2026-02-15T22:23:26+00:00","versionOfRecord":[],"versionCreatedAt":"2026-01-29 13:29:01","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-8605471","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-8605471","identity":"rs-8605471","version":["v1"]},"buildId":"XKTyCvWXoU3ODBz1xrDgd","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}