Continuous manufacturing of biologics is reshaping how the biopharmaceutical industry produces monoclonal antibodies, recombinant proteins, and advanced therapies. Unlike traditional fed-batch processing, where each unit operation runs as a discrete campaign with hold steps between stages, continuous bioprocessing connects perfusion culture directly to an integrated downstream train that operates without interruption. The result is a smaller facility footprint, higher volumetric productivity, and significantly lower cost of goods. Industry adoption has accelerated since the publication of ICH Q13 in 2023, and multiple commercial-stage integrated continuous biomanufacturing platforms now demonstrate that end-to-end continuous downstream processing of mAb products is technically and economically viable at production scale.
This guide covers every unit operation in the continuous manufacturing biologics workflow. It explains how perfusion bioreactors, periodic counter-current chromatography (PCC), continuous viral inactivation, and single-pass tangential flow filtration (SPTFF) connect into an integrated process. A detailed economic comparison and worked example quantify the COGS advantage for a 200 kg/year monoclonal antibody, and the regulatory and practical challenges are addressed honestly.
What Is End-to-End Continuous Manufacturing of Biologics?
End-to-end continuous manufacturing of biologics connects every unit operation in the production process so that material flows from the bioreactor through purification to formulated bulk drug substance without batch hold steps. In traditional manufacturing, a fed-batch culture is harvested, held in a tank, and processed through each downstream step as a discrete batch. Continuous bioprocessing eliminates these holds, running each operation at steady state and passing product continuously between steps.
The maturity of continuous manufacturing biologics implementations varies across the industry. Three levels describe the progression from partial to full integration:
Level 1: Continuous Unit Operations
Individual unit operations run continuously (e.g., perfusion bioreactor or PCC capture), but batch hold steps exist between stages. Most current commercial implementations operate at this level.
Level 2: Connected Sections
Upstream and downstream sections are internally connected (e.g., perfusion feeding directly into PCC, or continuous polishing through SPTFF), but the two sections are decoupled by a surge tank.
Level 3: Fully Integrated
All unit operations from bioreactor to formulated bulk operate as a single integrated continuous biomanufacturing process with PAT-driven control and real-time release testing.
The fully integrated continuous bioprocessing approach (Level 3) has been demonstrated at pilot and clinical scale by several groups. Warikoo et al. (2012) described the first end-to-end continuous bioprocessing platform producing recombinant therapeutic proteins, connecting a perfusion bioreactor to PCC capture, continuous viral inactivation, and continuous polishing. Steinebach et al. (2017) designed and operated a continuous integrated mAb production process with automated control across all unit operations. These demonstrations established the technical foundation that is now driving commercial adoption of integrated continuous biomanufacturing.
The Continuous Upstream Process: Perfusion Bioreactors
The continuous manufacturing biologics process begins with a perfusion bioreactor. Unlike fed-batch culture, where cells grow for 10–14 days in a closed vessel before harvest, perfusion culture continuously removes spent medium containing product while retaining cells inside the bioreactor using a cell retention device. This enables the culture to reach and maintain steady-state viable cell densities (VCD) of 40–80 × 106 cells/mL, compared to typical fed-batch peak densities of 10–25 × 106 cells/mL.
Two cell retention technologies dominate continuous bioprocessing upstream operations. Alternating tangential flow filtration (ATF) uses a diaphragm pump to cycle cell suspension back and forth across a hollow-fiber filter, generating high wall shear that prevents fouling while retaining cells. Tangential flow filtration (TFF) recirculates cell suspension across the filter surface using a pump. Both deliver cell-free harvest containing the product of interest at rates of 1–2 vessel volumes per day (VVD).
The productivity advantage of perfusion culture for continuous manufacturing of biologics is substantial. A 500L perfusion bioreactor operating at steady state with a VCD of 50 × 106 cells/mL, specific productivity of 20 pg/cell/day, and a perfusion rate of 1.5 VVD produces approximately 500 g of mAb per day. This translates to a volumetric productivity of roughly 1.0 g/L/day in the bioreactor and 2.0–2.5 g/L of purified product per day when accounting for downstream recovery. By comparison, a 10,000L fed-batch bioreactor producing 5 g/L over 14 days achieves approximately 0.36 g/L/day. The perfusion system thus delivers 3–5× higher volumetric productivity using a bioreactor that is 20 times smaller.
Continuous perfusion culture also provides a more consistent feed stream for downstream processing. The harvest composition (titer, host cell protein levels, aggregate content) stabilizes once the culture reaches steady state, typically within 5–7 days of initiating perfusion. This steady-state operation simplifies downstream chromatography loading and improves overall process robustness compared to the variable harvest compositions seen across a fed-batch campaign.
Perfusion Calculator
Model perfusion bioreactor performance: cell density, harvest rate, bleed strategy, and volumetric productivity.
Continuous Capture: Periodic Counter-Current Chromatography (PCC)
The continuous harvest stream from a perfusion bioreactor feeds directly into the capture step. In batch processing, a single Protein A column is loaded, washed, eluted, and regenerated before the next batch can be processed. Periodic counter-current chromatography (PCC) uses 2–4 columns operating in a staggered cycle to accept continuous feed without interruption, making it the workhorse of continuous downstream processing for mAb purification.
In a 3-column PCC configuration, one column loads while a second column captures breakthrough from the first, and a third column undergoes wash, elution, and regeneration. When the first column reaches its target loading, the roles rotate: the second column becomes the primary loading column, the third captures its breakthrough, and the first regenerates. This cycling eliminates the column idle time that wastes 40–60% of batch chromatography capacity.
The performance advantages of PCC for continuous bioprocessing are well documented. Resin utilization increases by 40–60% because each column is loaded to near-breakthrough rather than to a conservative safety margin. Buffer consumption drops by 30–50% because the smaller columns require proportionally less wash and equilibration buffer. Product recovery of 83–92% is consistently reported, matching or exceeding single-column batch performance. Steinebach et al. (2017) demonstrated a 3-column PCC process integrated with continuous upstream and downstream operations, achieving greater than 90% recovery with Protein A resin lifetimes exceeding 200 cycles.
The continuous capture eluate is a semi-continuous stream of low-pH product pools. Each column elutes sequentially, producing discrete eluate fractions that can be pooled in a small surge vessel before feeding the next downstream step. This semi-continuous eluate stream is well-suited to direct feed into continuous viral inactivation, as the low pH of the Protein A elution (pH 3.0–3.5) is close to the target pH for viral inactivation. For a deeper discussion of continuous chromatography modes including PCC, SMB, and MCSGP, see our dedicated guide.
Chromatography Calculator
Size chromatography columns, calculate resin requirements, and model loading, wash, and elution cycles.
Continuous Viral Inactivation
Viral inactivation by low pH treatment is a mandatory step in monoclonal antibody manufacturing. In batch processing, the Protein A eluate pool is adjusted to pH 3.5–3.7 and held for 60–90 minutes in a stirred tank before neutralization. Translating this to continuous bioprocessing requires ensuring that every molecule of product experiences the minimum required residence time at the target pH, a challenge that introduces the concept of residence time distribution (RTD) control.
Continuous viral inactivation for integrated continuous biomanufacturing uses a plug-flow reactor (PFR) rather than a mixed tank. In a PFR, fluid moves through a narrow tube or coiled channel with minimal axial mixing, so all molecules experience nearly identical residence times. The Protein A eluate, already at or near pH 3.5, enters the PFR and traverses its length over a minimum of 60 minutes before exiting for neutralization. The narrow RTD of a well-designed PFR ensures that no product shortcircuits the minimum hold time. This is critical for viral safety, as regulators require demonstration that 100% of the product experiences the validated inactivation conditions.
Dean vortex mixing in coiled tubular reactors provides secondary flow patterns that maintain radial mixing while preserving the narrow plug-flow RTD. This prevents channeling and dead zones that could compromise viral safety. Gomis-Fons et al. (2020) demonstrated model-based design and control of a continuous viral inactivation step as part of an integrated continuous mAb platform, confirming that plug-flow reactors can deliver the residence time uniformity required for regulatory acceptance.
After the PFR, the low-pH stream is neutralized by inline addition of base (typically Tris buffer at pH 7–8) using a static mixer. The neutralized stream may pass through a small surge vessel before entering the polishing step, or it can feed directly into an AEX column in flowthrough mode.
Continuous Polishing and Concentration
Following viral inactivation and neutralization, the continuous manufacturing biologics process moves into polishing and concentration. These steps remove residual host cell proteins (HCP), DNA, leached Protein A, and aggregates while concentrating the product to its final formulation strength.
Anion exchange (AEX) flowthrough chromatography is the most common polishing step in continuous downstream processing of mAb products. The neutralized viral inactivation output, at approximately pH 7–8, is loaded onto an AEX column or membrane adsorber. At this pH, mAb molecules (with isoelectric points typically above 7.5) carry a net positive charge and flow through the AEX bed without binding, while negatively charged impurities (DNA, HCP, endotoxin, leached Protein A) bind to the resin. AEX flowthrough is inherently suited to continuous operation because loading can proceed until the column approaches impurity breakthrough, at which point a second column takes over while the first regenerates.
Single-pass tangential flow filtration (SPTFF) replaces the batch UF/DF (ultrafiltration/diafiltration) step used in conventional processing. In SPTFF, the dilute product stream passes through a series of TFF cassettes connected in a serpentine path, concentrating the product by 2–4× in a single pass without recirculation. This eliminates the batch UF recirculation loop and its associated hold time. SPTFF is well-matched to the relatively dilute and steady-flow output of continuous upstream and capture steps. Concentration factors of 2–4× per SPTFF stage are typical. Multiple stages can be connected in series for higher concentration ratios.
Inline buffer exchange is accomplished using either a second SPTFF stage operating in diafiltration mode or by inline dilution followed by a further SPTFF concentration step. This converts the product from elution buffer conditions to its final formulation buffer without a discrete batch UF/DF operation. The continuous buffer exchange step can be monitored by inline conductivity and pH probes, providing real-time confirmation that the buffer composition meets the target specification.
The final step before fill-finish is sterile filtration through a 0.2 micrometre membrane. In a fully integrated continuous biomanufacturing process, this filtered bulk drug substance can feed directly into a continuous fill-finish line, though most current implementations include a sterile hold vessel between filtration and filling.
How Much Does Continuous Manufacturing Reduce COGS?
The economic case for continuous manufacturing of biologics centres on four cost drivers: facility capital, labour, materials, and quality control. Pollock et al. (2017) published the most widely cited techno-economic analysis, comparing integrated continuous bioprocessing against traditional fed-batch manufacturing for clinical and commercial antibody production. Their analysis showed COGS reductions of 30–50% for continuous processes at commercial scale, depending on production volume and the degree of integration.
The COGS advantage of continuous bioprocessing emerges from several converging factors. Smaller bioreactors (500L vs 10,000L) reduce both capital cost and facility footprint. Higher resin utilization in PCC reduces Protein A costs, which are typically the single largest materials expense. Continuous operation requires fewer operators per kilogram of product. And the smaller facility footprint reduces cleanroom HVAC and utilities costs.
For a high-demand monoclonal antibody requiring 200 kg/year, the COGS differential between continuous and fed-batch manufacturing is striking. Traditional fed-batch at $130/g translates to $26 million/year in manufacturing costs. Intensified fed-batch (using N-1 perfusion seed and a smaller production bioreactor) reduces this to $95/g or $19 million/year. Fully continuous bioprocessing at $63/g brings the annual cost down to $12.6 million, saving $13.4 million per year compared to traditional fed-batch. Rathore et al. (2025) provide a critical examination of whether end-to-end continuous bioprocessing is always necessary, noting that the economic advantage depends heavily on scale and product demand.
The facility footprint advantage is equally dramatic. A traditional fed-batch facility producing 200 kg/year requires two 10,000L stainless-steel bioreactors plus correspondingly large downstream equipment, housed in a facility of approximately 3,000–4,000 m2 of cleanroom space. An equivalent continuous manufacturing biologics facility using a 500L perfusion bioreactor with integrated continuous downstream processing fits into 1,000–1,500 m2, a 50–70% footprint reduction. This translates to substantially lower facility construction costs and faster time to first GMP batch.
| Unit Operation | Batch Mode | Continuous Mode | Improvement |
|---|---|---|---|
| Upstream culture | 10,000L fed-batch, 5 g/L in 14 days (0.36 g/L/day) | 500L perfusion, 50 × 106 cells/mL, 1–2 VVD (2.5 g/L/day) | 3–5× volumetric productivity; 20× smaller vessel |
| Protein A capture | Single column, 30–40 g/L resin loading, 40–50% utilization | 3-column PCC, 50–60 g/L loading, 80–95% utilization | 40–60% higher resin utilization; 30–50% less buffer |
| Viral inactivation | Batch hold in stirred tank, 60–90 min, pH 3.5 | Plug-flow reactor, ≥60 min residence time, narrow RTD | Continuous flow; smaller vessel; tighter RTD control |
| Polishing | Single AEX column, batch load-wash-strip | Twin-column AEX flowthrough, alternating load/regen | Continuous operation; higher throughput per resin volume |
| UF/DF concentration | Recirculating TFF, 4–8 h batch operation | SPTFF, 2–4× concentration in single pass | Eliminates recirculation loop; no batch hold |
| Overall process | 15-day campaign cycle, multiple hold steps, large facility | Steady-state operation, 300+ days/year, 50–70% smaller footprint | 30–50% COGS reduction; 50–70% footprint reduction |
Worked Example: 200 kg/year mAb Production Comparison
Scenario: A commercial monoclonal antibody requires 200 kg/year of purified drug substance. Compare a traditional fed-batch facility against a fully continuous bioprocessing platform.
Option A: Traditional Fed-Batch
- 2 × 10,000L stainless-steel bioreactors (8,000L working volume)
- Titer: 5 g/L over 14-day culture → 8,000L × 5 g/L = 40 kg per batch (harvest)
- DSP recovery: ~70% → 40 kg × 0.70 = 28 kg purified per batch
- Turnaround: 18 days per batch (14-day culture + 2-day harvest/CIP + 2-day turnaround)
- Batches needed: 200 kg / 28 kg = ~7 batches/year (4 from reactor A, 3 from reactor B)
Annual purified output = 7 batches × 28 kg/batch = 196 kg ≈ 200 kg
COGS = ~$130/g × 200,000 g = $26,000,000/year
Option B: Fully Continuous
- 1 × 500L perfusion bioreactor (steady-state 50 × 106 cells/mL)
- Specific productivity: 20 pg/cell/day at 1.5 VVD = ~500 g mAb/day in harvest
- DSP recovery: ~80% overall (PCC 90% × VI 98% × AEX 95% × SPTFF 95%)
- Operating days: 300 days/year (accounting for CIP, turnarounds, and planned maintenance)
Daily harvest = 50 × 106 cells/mL × 20 pg/cell/day × 500 L = 500 g/day
Daily purified output = 500 g/day × 0.80 DSP yield ≈ 400 g/day
But: ~33 pg/cell/day qP with optimized clones lifts harvest to ~825 g/day × 0.80 = 660 g/day
Annual output = 667 g/day × 300 days = ~200 kg/year
COGS = ~$63/g × 200,000 g = $12,600,000/year
Result: The continuous bioprocessing platform saves approximately $13.4 million per year (52% COGS reduction) while using a bioreactor 20× smaller and a facility footprint 50–70% smaller than the fed-batch alternative. The continuous process requires fewer operators (3–4 vs. 8–10 per shift) and produces more consistent product quality due to steady-state operation.
Regulatory Considerations for Continuous Biomanufacturing
The regulatory landscape for continuous manufacturing of biologics has matured significantly. ICH Q13, "Continuous Manufacturing of Drug Substances and Drug Products," was finalized in November 2023 and provides the harmonized framework that FDA, EMA, and PMDA now reference for continuous bioprocessing submissions. The guideline addresses the unique aspects of continuous manufacturing that differ from batch processing, including batch definition, system dynamics and control strategy, process monitoring, material traceability, and process validation.
Batch definition is a fundamental regulatory question for continuous biomanufacturing. In batch processing, a batch is defined by the contents of the bioreactor at harvest. For a continuous process running for months, regulators require a clear definition of what constitutes a batch for release testing, stability studies, and traceability. ICH Q13 allows batch boundaries to be defined by time (e.g., one batch equals 24 hours of continuous output), by quantity (e.g., one batch equals the amount needed for a defined number of drug product lots), or by processing events (e.g., one batch equals the material produced between planned shutdowns).
Process Analytical Technology (PAT) is essential for continuous bioprocessing. The FDA PAT framework, originally published in 2004, encourages real-time monitoring and control as the basis for process understanding. In continuous manufacturing biologics, PAT sensors at critical control points provide the data needed to confirm that product quality attributes are met continuously rather than tested only at batch release. Raman spectroscopy monitors glucose, lactate, and product concentration in the bioreactor. UV 280 nm detectors at chromatography outlets track product elution and breakthrough. Inline pH and conductivity sensors verify viral inactivation conditions and buffer exchange endpoints.
Process validation for continuous processes follows the same lifecycle approach (Stage 1: Process Design, Stage 2: Process Qualification, Stage 3: Continued Process Verification) as batch processes, but with modifications. Stage 2 qualification must demonstrate that the process achieves steady-state performance, that startup and shutdown transitions produce acceptable product, and that the process can handle planned and unplanned disturbances (e.g., column switching in PCC, surge tank level fluctuations, PAT sensor replacement). Continued process verification (Stage 3) is particularly important for continuous manufacturing, as the continuous data stream enables ongoing statistical monitoring of process performance.
Real-time release testing (RTRT) is a natural fit for integrated continuous biomanufacturing. By correlating inline PAT measurements with offline analytical results during process development and qualification, manufacturers can establish validated models that predict final product quality from process data alone. This can reduce or eliminate end-of-batch release testing for certain quality attributes, shortening the time from production to patient and reducing analytical costs.
Challenges and Limitations
Despite the compelling economic and technical advantages, continuous manufacturing of biologics faces practical challenges that explain why most of the industry still operates in batch or semi-continuous modes.
Surge tank management is a critical design challenge for end-to-end continuous bioprocessing. The upstream perfusion bioreactor produces harvest continuously, but downstream unit operations (particularly PCC elution) produce output in semi-continuous pulses. Surge tanks between unit operations absorb these flow rate mismatches, but they add complexity, must be sized correctly, and introduce potential hold-time constraints that can affect product quality. Undersized surge tanks force the process to stop when downstream operations slow. Oversized surge tanks negate the footprint advantages of continuous processing.
PAT integration complexity increases with the number of connected unit operations. Each PAT sensor requires qualification, calibration maintenance, and data management infrastructure. The control system must interpret data from multiple sensors simultaneously and make real-time decisions about process parameters, column switching, flow rates, and diversion of out-of-specification material. This demands a higher level of automation sophistication than batch processing, where operators can intervene between discrete steps.
Process validation burden for fully integrated continuous biomanufacturing is heavier than for batch processes. In addition to demonstrating steady-state performance, manufacturers must validate startup, shutdown, and transition phases. They must show that the process responds appropriately to disturbances such as a PAT sensor failure, an air bubble in the feed line, or a temporary loss of one PCC column. The validation strategy must address material traceability across continuously connected unit operations, which is more complex than tracking discrete batch lots.
Operator training and organizational readiness represent underestimated barriers to continuous bioprocessing adoption. Operating a continuous manufacturing biologics process requires different skills from batch manufacturing. Operators must understand process dynamics, interpret real-time PAT data, and make rapid decisions during transient events. Manufacturing organizations accustomed to batch processing need significant investment in training, procedures, and culture change before they can operate continuous processes reliably.
Economic crossover point: Continuous manufacturing is not always the most cost-effective choice. For low-volume products (below roughly 100 kg/year), niche indications, or products with uncertain commercial demand, the additional process development investment and operational complexity of continuous bioprocessing may not be justified. Single-use fed-batch bioreactors with disposable downstream flowpaths can offer lower COGS for small-scale production with minimal capital investment, as explored in our bioreactor selection guide. Rathore et al. (2025) argue that hybrid approaches, combining continuous upstream with batch downstream or vice versa, may offer the best balance of cost and complexity for many products.
Frequently Asked Questions
What is end-to-end continuous manufacturing of biologics?
End-to-end continuous manufacturing of biologics is an integrated production approach where all unit operations, from cell culture through final formulation, operate continuously without batch hold steps. A perfusion bioreactor feeds a connected downstream train of continuous capture (PCC), viral inactivation (plug-flow reactor), polishing (AEX flowthrough), and concentration (SPTFF), achieving 30–50% lower COGS and 50–70% smaller facility footprint compared to traditional batch processing.
How much does continuous bioprocessing reduce manufacturing costs?
Continuous bioprocessing typically reduces cost of goods sold (COGS) by 30–50% compared to traditional fed-batch manufacturing. For a 200 kg/year monoclonal antibody, COGS drops from approximately $130/g with dual 10,000L fed-batch bioreactors to roughly $63/g with a 500L perfusion system and integrated continuous downstream processing. The savings come primarily from reduced facility costs (smaller equipment and footprint), lower labor requirements, higher resin utilization, and reduced buffer consumption.
What regulatory guidance applies to continuous biomanufacturing?
ICH Q13, finalized in 2023, provides the primary regulatory framework for continuous manufacturing of drug substances and drug products. It covers batch definition for continuous processes, process monitoring and control strategy, real-time release testing, and process validation approaches. The FDA PAT (Process Analytical Technology) framework and ICH Q8–Q12 guidelines also support continuous bioprocessing by encouraging science-based and risk-based approaches to process design and control.
What is PCC chromatography and why is it used in continuous bioprocessing?
Periodic counter-current chromatography (PCC) uses 2–4 columns in a staggered loading cycle where breakthrough from one column is captured by the next, achieving 40–60% higher resin utilization and 30–50% lower buffer consumption compared to single-column batch chromatography. PCC is the dominant continuous capture format for Protein A purification of monoclonal antibodies, delivering 83–92% mAb recovery with consistent product quality across cyclic steady state.
Is continuous manufacturing always better than fed-batch for biologics?
No. Continuous manufacturing biologics offers the greatest advantage at high production volumes (above roughly 100 kg/year), where the capital and operating cost savings justify the additional process development complexity. For low-volume specialty biologics, niche indications, or early clinical supply, fed-batch manufacturing with single-use bioreactors may be more cost-effective and operationally simpler. The decision depends on production scale, product lifetime demand, facility strategy, and organizational readiness for continuous operations.
Related Tools
- Perfusion Calculator — Model perfusion bioreactor performance including cell density, harvest rate, and bleed strategy.
- Chromatography Calculator — Size chromatography columns, calculate resin requirements, and model loading and elution.
- Scale-Up Calculator — Scale bioprocess parameters between vessel sizes using geometric and engineering criteria.
- Filtration Calculator — Size TFF and normal flow filtration systems for UF/DF and sterile filtration.
- Fermentation Economics — Compare COGS across manufacturing scales and process configurations.
References
- Pollock J, Coffman J, Ho SV, Farid SS. Integrated continuous bioprocessing: Economic, operational, and environmental feasibility for clinical and commercial antibody manufacture. Biotechnol Prog. 2017;33(4):854–866. doi:10.1002/btpr.2492
- Warikoo V, Godawat R, Brower K, Jain S, Cummings D, Simons E, Johnson T, Walther J, Yu M, Wright B, McLarty J, Karey KP, Hwang C, Zhou W, Riske F, Konstantinov K. Integrated continuous production of recombinant therapeutic proteins. Biotechnol Bioeng. 2012;109(12):3018–3029. doi:10.1002/bit.24584
- Steinebach F, Ulmer N, Wolf M, Decker L, Schneider V, Wälchli R, Karst D, Souquet J, Morbidelli M. Design and operation of a continuous integrated monoclonal antibody production process. Biotechnol Prog. 2017;33(5):1303–1313. doi:10.1002/btpr.2522
- Gomis-Fons J, Schwarz H, Zhang L, Andersson N, Nilsson B, Castan A, Solbrand A, Stevenson J, Chotteau V. Model-based design and control of a small-scale integrated continuous end-to-end mAb platform. Biotechnol Prog. 2020;36(4):e2995. doi:10.1002/btpr.2995
- Rathore AS, Metya S, Nitika. Do We Really Need End-To-End Continuous Processing for Biomanufacturing of Monoclonal Antibodies? Biotechnol Bioeng. 2025. doi:10.1002/bit.70147