By Toni Nikolova, Kurt Brorson, and David Perez, Parexel Regulatory Consulting Services
For over three months the world has been battling COVID-19. COVID-19 is unforgiving and has the ability to overwhelm even the strongest health system or regulatory regime. Even though the immediate measures put in place proved to be efficacious so far, the world is clearly transitioning into a new ‘normal’ and the pharma/biopharma world needs to transition too.
These very trying times emphasize even more the importance of manufacturing drugs that are not only safe but available to the general public expeditiously. Our Parexel colleagues have authored a number of blogs, describing the risk of COVID-19 transmission through biotech and other products, e.g., monoclonal antibodies (mAbs), cell and gene therapy as well as measures that can be taken to reduce virus risk, like filtration. In this blog, we discuss an idea that can expedite product development: modular validation.
The idea of modular virus clearance, i.e. applying data from one product to another, was first introduced in a forward-thinking regulatory document from FDA/CBER (Center for Biologics Evaluation and Research at FDA) who at the time oversaw therapeutic monoclonal antibodies: “Points to Consider in the Manufacture and Testing of Monoclonal Antibody Products for Human Use” (1997). While now considered withdrawn by FDA, the document promulgated specific applicability criteria for modular claims, but in summary the concept is “Different model mAb may be used to demonstrate viral clearance in different modules... and then identical modules in the procedure may be extrapolated to the product mAb”. The document clarifies that “Modular clearance will apply only when the mAbs have similar biochemical properties and are purified by identical methods”. It also advises that “Particular attention should be paid to column elution, buffer conditions, including pH and ionic strength, the sequence of columns, protein concentration, dwell times, flow rate, pressure, temperature, and potential problems associated with scale-up.” These are all standard scale-down factors that biotech manufacturers consider when they develop and qualify small scale models of their process.
Many biotech companies have several products in development with an already studied virus removal unit operation. These can include column steps like protein A, anion exchange columns or hydrophobic interaction columns. They may also include specific viral clearance steps like virus filters or inactivation steps. To make a modular claim, it is important to understand how the unit operations remove viruses. For example, consider capture chromatography. Viruses, including retrovirus-like particles (RVLP) present in all mammalian cell culture harvests, mostly flow through a Protein A column during loading, along with the most non-antibody cell culture components. However, a small number of RVLP do bind to the column via non-specific interaction, a phenomenon that is not well understood, with the media or the mAb. These viruses are dislodged during the pH transition of product elution, thus leading to variable levels of residual virus in the eluate. Studies have shown that this effect can depend on the specific product-containing harvest load material. Thus, given this complexity and uncertainty, RVLP removal by protein A columns isn’t generally regarded to be robust and reliable.
On the other hand, low pH incubation has been shown in multiple studies to a robust, reliable and effective step to inactivate many non-enveloped viruses because the low pH deprotonates and denatures their external membrane proteins. Due to this robustness, an ASTM standard (ASTM. 2019. E2888-12, Standard Practice for Process for Inactivation of Rodent Retrovirus by pH. ASTM International, Conshohocken PA) for low pH inactivation (E2888-12) has been established. Namely, under specific conditions of temperature, time, pH, buffer species and protein, a 5 log10 inactivation of murine retroviruses can be assumed.
Virus retentive filters, used towards the end of bioprocess trains when the process fluids are largely pure product protein, remove viruses from product streams by a sieving based mechanism. Virus filters are designed to target either larger viruses (e.g., retroviruses) or small viruses (e.g., parvoviruses). Removal of large viruses like murine retroviruses by both large and small virus retentive filters has been shown in multiple studies to be highly robust and effective, and subject matter experts consider that modular claims are possible. This is because retroviruses are 80-110 nm, while large retentive filter pores are around 35-50 nm and the parvo filter pore size is around 20 nm. In the case of small virus removal, the scientific literature has identified some failure modes to be avoided. Thus, modular approaches can also be contemplated for small virus removal but must mitigate the risks posed by the failure modes by ensuring a steady transmembrane pressure, avoidance of filter overloading as well as controlling feed-stream purity and protein concentration.
The idea of the modular viral clearance has been also embraced by EMA This concept was first introduced by EMA in the “Guideline on Virus Safety Evaluation of Biotechnological Investigational Medicinal Products” (EMEA/CHMP/BWP/398498/2005) although using a different terminology related to “prior in-house knowledge” of the manufacturing process viral clearance. “In the event that a manufacturer is developing similar types of products by established and well-characterized procedures, virus reduction data derived for these other products might be applicable to the new product for an equivalent processing step. In general, in order to make use of data from such a step, the step should have been carefully evaluated, including a thorough study of the process parameters that affect virus reduction. If data for more than one product is available for the specific step, the effectiveness of virus reduction should be comparable in each case”. The manufacturer needs to provide a rationale of why prior in-house data can be applied to a new product, e.g. product intermediate at the stage before such a step has comparable biochemical properties and is purified by identical methods, if new products have components not present in previous products, their potential impact in viral clearance needs to be determined. The analysis should provide complete confidence in the conclusion that in both cases the established manufacturing step is similar in its capacity to inactivate/remove potential virus contaminants.
Given the current environment, regulatory flexibility is warranted to streamline development of anti-COVID-19 products. Application of modular validation during early product development is one potential avenue for streamlining by postponing costly and time-consuming viral clearance studies until a phase in development when the utility of the new molecular entity (NME) is clearer. This is a huge advantage for, as an example, newly developed monoclonal antibodies against COVID-19 envelope proteins, where time is of the essence to transition them into the clinic. Understanding viral clearance by their specific manufacturing processes is important, but if platform clearance by robust, well studied steps can be leveraged initially, the NME to clinic time can be reduced by at least a couple months.
Parexel’s consulting group can provide clients with expert advice when developing comprehensive virus risk mitigation strategies for biopharmaceuticals and biologics. Our recommendations are based on a deep knowledge and experience in viral clearance in bioprocessing. As mentioned above, regulators and the biologics industry are confronted with and have developed risk mitigation strategies against emerging viruses, including modular viral clearance. Parexel stands ready to assist companies to develop tailored virus risk-mitigation strategies that are based on deep expertise.