Addressing Protein Aggregation

Biopharmaceuticals possess the capability to treat severe illnesses and can generally be considered as relatively safe, but despite stable formulations resulting in high product quality, the formation of protein aggregates can occur for a number of reasons, including chemical or physical degradation, such as oxidation or denaturation (1). Aggregates can be small and soluble or grow into larger particles. Other factors that contribute to particle generation include light, shearing, shaking or temperature (2)(3)(4). The majority of protein drug products will encounter most of the above mentioned factors at some point during their production cycle and shelf life – and the risk associated with these protein aggregates is that they can endanger drug safety and efficacy (5)(6). Immune reactions caused by non-native protein species have been known about since the 1950s (7). Despite many improvements in the generation of recombinant proteins, such as fully humanized proteins or sequence modifications, it is still rather common to detect the formation of anti-drug antibodies in the blood of patients treated with protein drug products because of protein aggregation (8)(9). 

In the majority of cases, anti-drug antibodies only have minor clinical relevance, but severe events – anaphylaxis, serum sickness or life-threatening cases like the neutralization of an endogenous protein – can occur (5)(10). Furthermore, beside proteinaceous particles, a protein drug product can contain other particles, such as silicone oil, glass microflakes, rubber, plastic or metal (11)(12)(13). Combined particles, such as non-proteinaceous particles covered with native or non-native protein species, can also be formed. Some of these combinations, like protein adsorbed to silicone oil, are known to trigger an immune response (14). However, with the immunogenic potential of each of the possible particle subgroups, one could theoretically encounter in a protein drug product that is not yet known, and unlikely to be clarified in the near future.

Although the correlation between immunogenicity and protein particles is commonly accepted, plenty of other factors, such as the immune status of the patient, dose, dosing frequency or route of administration, also play a role (7)(15)(16)(17). Handling, transportation and storage of the drug after the drug product release by the manufacturer can also impact product quality (18).

The problem with packaging

It is important for biopharmaceutical manufacturers to consider how primary packaging materials impact product stability, particularly for sensitive biopharmaceuticals. Protein drug products are generally filled either in vials, syringes or cartridges made of borosilicate glass (18). For glass syringes and cartridges, silicone oil is necessary to enable smooth gliding and low break-loose forces for injection. The problem with silicone oil is that it can contribute to the particle burden of the product by shedding from the glass barrel and is also a known agent that fosters the formation of protein aggregates (13)(19). Silicone oil microdroplets derived from prefilled syringes can also act as an adjuvant leading to induction of anti-drug antibodies (14). Although multiple techniques, such as baked on or cross-linked siliconization, exist to reduce the amount of free silicone oil in glass syringes (19)(20), interactions still occur between the silicone oil and the protein solution. 

Eliminating silicone oil could be seen as beneficial, but alternative materials would be needed. One option is to use plastic syringe barrels made of cyclic olefin (co-) polymers. Several major syringe manufacturers have these polymer-based syringes in their portfolio, but only two systems are completely free of silicone oil (21) – both use new coating technologies for the stoppers that enable functionality without silicone oil. For the storage of biopharmaceuticals, this is a major advantage and it has been shown that the particulate burden of a solution can drastically be reduced in silicone oil free polymer syringes (18). Further, these syringes can be produced without tungsten and glue in the case of staked needles, eliminating other potential complication (22)(23)(24). A major shortcoming of these syringes might be their higher oxygen permeability in comparison to glass (25), but simple modifications may overcome the problem; for example, designing syringes with multiple layers that possess higher gas barrier properties. Cheaper solutions include using oxygen-tight labels or sealing the syringe in a gas-tight aluminum bag (26). So far, these syringes have not been evaluated for the long-term storage of oxygen-sensitive biopharmaceuticals, but we are currently investigating this topic. For the moment, we can confirm other reports about lower particle counts in polymer-based syringes. Oxygen permeability is also controllable with easy modifications, as our study has shown.

Increasing patient safety

Although improvements on the primary packaging side may lead to better products, no one can guarantee that the quality of every single drug container, particularly in regards to the overall particle burden and the nature of the aggregate type, is the same. We would like to propose an approach that should be easy to implement and that has the capability to reduce a potential risk for patients from particulate matter to a minimum.

Our concept is based on an expansion of already used bedside filtration to a much broader range of products. This should provide increased safety to every single container. To support our idea, we carried out a survey analysis on more than 300 marketed protein drug products. We found that 16 percent of them are already filtered during bedside preparation and administration of the drug. Only a handful of drugs had explicit statements not to use filtration (27), so there is great potential to expand bedside filtration. Today, regulatory authorities require greater monitoring of particles in the low micrometer range – and more of the recently approved drugs are filtered. Our analysis also revealed that specific recommendations for filtration are rare and the user is often left alone at this point. If specific instructions were included, polyethersulfone membranes with a pore size of 0.2 µm were the most commonly used (27).

Using a filter with a pore size of 0.2 µm has several benefits. First, these filters are broadly available and commonly used in clinical settings, enabling easy, quick and cheap distribution. Second, these filters remove most particles above 0.2 µm as our data show. Third, another final sterile filtration is carried out. With that, the patient benefits from other effects of bedside filtration, including reduced occurrence of infection, sepsis or thrombi (28).

A change to more frequent usage of bedside filtration will not occur overnight. So, the question of how to establish common routine bedside filtration remains. A start would be to integrate the filtration step into processes where several handling steps are already necessary. For example, the preparation and administration of lyophilisates, include addition of the solvent, swirling the vial until complete dissolution of the powder, inspection of the vial for particulates, aspiration into a syringe, change of the needle, and finally administration. Other applications with several handling steps include multi-dose vials or mono-dose vials where the solution needs to be aspirated into the administration syringe. Also in the case of infusion, a filter proximal to the patient can be easily included (29).

Our work has shown that filtration has huge potential for eliminating protein aggregates for multiple relevant protein drug products, and we have not encountered problems such as protein adsorption or denaturation. Although our data show that filtration is easily possible, the filtration process has to be further improved with the help of the filter industry. New filter designs with lower hold up volumes and filtration areas are necessary to expand the concept of bedside filtration to smaller volumes. Filter cleanliness has to be assured by the filter manufacturers, which includes not only the particle burden of the filters themselves, but also extractable and leachable profiles, which are partly lacking (30). Based on the generated experimental data, which will be published soon, we believe that bedside filtration can contribute significantly to patient safety.

It is important to emphasize that the quality of today’s protein drug products is very high. Our proposal for an expansion of bedside filtration is not an attempt to cover any shortcomings in product development or manufacturing, but an immediate opportunity to further improve the safety and efficacy for patients. 
By Benjamin Patrick Werner, and Gerhard Winter Professor, both in the Department of Pharmacy, Pharmaceutical Technology & Biopharmaceutics, Ludwig-Maximilians-University, Munich, Germany.