2. The ICH Q5A Framework
ICH Q5A establishes three complementary pillars for demonstrating viral safety. All three must be addressed in a regulatory submission; no single pillar is sufficient on its own.
Pillar 1: Cell Line Characterization
The Master Cell Bank (MCB) and Working Cell Bank (WCB) must be tested for the presence of endogenous and adventitious viruses. For CHO cells, this includes:
- In vitro virus assays: Co-cultivation with multiple detector cell lines (MRC-5, Vero, CHO) to detect cytopathic, hemadsorbing, and hemagglutinating agents.
- In vivo assays: Inoculation into suckling and adult mice, guinea pigs, and embryonated eggs (being phased out in favor of molecular methods in ICH Q5A(R2)).
- TEM: Examination for retrovirus-like particles. CHO cells consistently show Type A (intracisternal) and Type C (budding) particles.
- Reverse transcriptase (RT) activity: Quantitative product-enhanced RT (qPERT) assay detects retroviral RT activity at high sensitivity.
- PCR-based testing: Broad-range PCR panels for specific viral families, and next-generation sequencing (NGS) for comprehensive viral detection.
Pillar 2: Testing of Production Materials
The unprocessed bulk harvest (cell culture supernatant before purification) must be tested for adventitious agents. This serves as a lot-release safety net to detect any viral contamination that may have occurred during production. Testing includes in vitro virus assays, PCR panels, and increasingly, NGS-based metagenomics.
Pillar 3: Viral Clearance Validation
This pillar demonstrates that the downstream purification process has the capacity to remove or inactivate viruses, even if a contamination event were to occur. Viral clearance studies are performed by spiking model viruses into scale-down process intermediates and measuring virus reduction across each purification step. The cumulative log reduction value (LRV) across the entire process must meet defined targets for both retroviral and non-enveloped model viruses.
3. Virus Reference Panel
Viral clearance studies use well-characterized model viruses that represent a range of physicochemical properties. The panel must include viruses that model the worst-case scenarios for the production cell line being used. For CHO-derived products, the standard panel includes:
| Model Virus | Genome | Envelope | Size (nm) | Models For |
|---|---|---|---|---|
| X-MuLV (Xenotropic MuLV) | RNA | Yes | 80–110 | Endogenous retrovirus |
| MMV (Minute virus of mice) | DNA | No | 18–24 | Small non-enveloped (hardest to clear) |
| PRV (Pseudorabies virus) | DNA | Yes | 120–200 | Large enveloped DNA virus |
| Reo-3 (Reovirus type 3) | RNA | No | 60–80 | Medium non-enveloped |
MMV (or its close relative MVM, minute virus of mice) is the most challenging virus to clear because it is small (18–24 nm), non-enveloped (resistant to low pH and detergent inactivation), and extremely stable. It passes through many chromatography steps and is only reliably removed by nanofiltration with 20 nm pore size filters. The ability to demonstrate adequate MMV clearance is often the limiting factor in a viral clearance strategy.
The choice of specific model viruses should be justified based on the cell substrate, raw materials used, and the geographic region of manufacturing. Some regulatory agencies may request additional viruses—for example, the use of bovine-origin raw materials may require testing with bovine parvovirus (BPV) or bovine viral diarrhea virus (BVDV).
4. Clearance Steps & Typical LRVs
Each downstream purification step contributes to overall viral clearance through either removal (partitioning) or inactivation mechanisms. The key distinction matters: a robust viral clearance strategy must include both inactivation and removal steps operating through orthogonal mechanisms.
| Process Step | Mechanism | LRV (Enveloped) | LRV (Non-enveloped) |
|---|---|---|---|
| Low pH hold (pH 3.5, 60 min) | Inactivation | >4 | 0–1 |
| Protein A chromatography | Partitioning | 1–3 | 1–2 |
| Anion exchange (AEX, flow-through) | Partitioning | 3–5 | 3–5 |
| Nanofiltration (20 nm) | Size exclusion | >4 | >4 |
| Solvent/detergent (S/D) | Inactivation | >4 | 0 |
| Cation exchange (CEX, bind/elute) | Partitioning | 2–4 | 2–4 |
Regulators expect viral clearance to be achieved through at least two orthogonal mechanisms. For example, low pH inactivation and nanofiltration (size-based removal) are orthogonal. Two chromatography steps operating on the same principle (e.g., two ion exchange steps) may not be accepted as fully independent clearance steps, even if each demonstrates measurable LRV individually.
The LRV Equation
The log reduction value quantifies the effectiveness of each step:
where:
V = volume of load or eluate/filtrate
T = virus titer (infectious units/mL)
Total process LRV is the sum of individual step LRVs, provided each step operates through an independent mechanism. Steps with LRV < 1.0 are generally not claimed as contributing clearance steps.
5. Building Your Strategy
The goal is to demonstrate sufficient cumulative LRV for both the retrovirus model (X-MuLV) and the small non-enveloped virus model (MMV). The generally accepted targets are:
Retrovirus (X-MuLV): ≥ 12 LRV
Small non-enveloped (MMV): ≥ 6 LRV
Example Platform Process
A typical monoclonal antibody purification process and its expected viral clearance:
This process meets both targets with comfortable margins. The strategy includes two inactivation-independent removal steps (AEX and nanofiltration) plus one inactivation step (low pH), providing orthogonality. Protein A partitioning is claimed but is not considered a dedicated viral clearance step—it contributes to the overall safety margin.
Calculate Your Viral Clearance
Enter your process steps and virus panel to calculate cumulative LRVs and check compliance against ICH Q5A targets.
Viral Clearance Calculator →Strategy for Products Without Low pH Hold
Some products (e.g., acid-labile fusion proteins, non-Protein A purified proteins) cannot use low pH inactivation. In these cases, alternative inactivation steps such as solvent/detergent treatment can be used for enveloped viruses. For non-enveloped virus clearance without low pH, the process relies more heavily on chromatographic partitioning and nanofiltration. Consider adding a second chromatography step or using larger nanofilters (15 nm) to increase the safety margin.
6. Common Pitfalls
Two ion exchange chromatography steps operating at similar conditions (same resin chemistry, similar pH and conductivity) may not be accepted as independent clearance steps by regulators. The clearance mechanism is the same—electrostatic partitioning—so the LRVs are not additive in the eyes of the agency. Use steps with genuinely different clearance mechanisms: size exclusion (nanofiltration), chemical inactivation (low pH, S/D), and charge-based partitioning (ion exchange).
The virus spike (typically 5–10% v/v of crude virus stock added to the process intermediate) can alter the composition of the load material. High-titer virus stocks prepared in cell culture contain host cell proteins, DNA, and media components that can affect chromatography performance. Always verify that the spiked load behaves comparably to unspiked material by monitoring protein recovery, step yield, and product quality attributes. Include a “hold control” (spiked material held at process temperature for the duration of the step) to distinguish inactivation from removal.
Process intermediates (low pH fractions, high-salt fractions, detergent-containing solutions) can be cytotoxic to the indicator cells used in the virus infectivity assay. This cytotoxicity limits the assay sensitivity, reducing the maximum detectable LRV. If your post-step sample can only be diluted 1:10 before assaying, and the input titer was 106, the maximum demonstrable LRV is only 5 (not the actual clearance, which may be higher). Pre-study cytotoxicity assessments and sample dilution strategies are essential.
Viral clearance studies are always performed at laboratory scale using scale-down models of the manufacturing process. These models must be qualified to demonstrate comparability with the full-scale process. Key parameters to match include linear flow rate, column bed height, load ratio, buffer compositions, and product quality. Document the scale-down model qualification in the viral clearance study report—regulators will specifically review this section.
7. Regulatory Submission
Viral clearance data is included in Module 3.2.A.2 of the Common Technical Document (CTD) for BLA (US) or MAA (EU) submissions, under the section titled “Adventitious Agents Safety Evaluation.” The submission must include:
- Study design and rationale: Justification for model virus selection, description of scale-down model, and process step selection for evaluation.
- Virus characterization: Identity, titer, and purity of virus stocks used for spiking.
- Scale-down model qualification: Demonstration that the lab-scale model is representative of manufacturing scale.
- Raw data: Virus titers at each step (load, eluate/filtrate, wash fractions), including all replicates and assay controls.
- Statistical analysis: Calculation of LRV with confidence intervals. When virus is not detected in the output, the LRV is reported as “≥” the limit of detection.
- Overall viral clearance evaluation: Cumulative LRV table across all evaluated steps, with assessment of orthogonality and comparison to acceptance targets.
For related downstream process calculations, see our Chromatography Calculator for column sizing and buffer volume estimates, and the Filtration Calculator for nanofiltration membrane area requirements.
Begin viral clearance studies during late Phase 2 or early Phase 3 development, once the purification process is locked. Studies take 3–6 months to complete through a contract testing laboratory (e.g., Texcell, BioReliance, Charles River). Allow additional time for report writing and regulatory review. Late initiation of viral clearance studies is one of the most common causes of BLA submission delays.
References
- ICH Q5A(R2). “Viral Safety Evaluation of Biotechnology Products Derived from Cell Lines of Human or Animal Origin.” International Council for Harmonisation, 2024. ICH Q5A(R2)
- Brorson, K., Krejci, S., Lee, K., Hamilton, E., Stein, K., & Xu, Y. (2003). “Bracketed generic inactivation of rodent retroviruses by low pH treatment for monoclonal antibodies and recombinant proteins.” Biotechnology and Bioengineering, 82(3), 321–329. doi:10.1002/bit.10574
- Miesegaes, G., Lute, S., & Brorson, K. (2010). “Analysis of viral clearance unit operations for monoclonal antibodies.” Biotechnology and Bioengineering, 106(2), 238–246. doi:10.1002/bit.22662
- PDA Technical Report No. 41 (Revised 2013). “Virus Filtration.” Parenteral Drug Association.
- Strauss, D.M., Cano, T., Cai, N., et al. (2009). “Strategies for developing design space for viral clearance by anion exchange chromatography during monoclonal antibody production.” Biotechnology Progress, 25(6), 1754–1764. doi:10.1002/btpr.263