The availability of automated methods for the compounding of monoclonal antibodies provides the opportunity to reduce valuable spillage and to reproducibly compound a high-quality pharmaceutical product
Bas JM Peters PhD PharmD
Ewoudt MW van de Garde PhD PharmD
Department of Clinical Pharmacy, St Antonius Hospital, Nieuwegein, The Netherlands
Email: [email protected]
A growing number of monoclonal antibodies (mAbs) are available on the market and many more are expected to follow.(1) mAbs are proteins and are available as a liquid or lyophilised formulation. mAbs need to be compounded in order to be administrated intravenously. Compounding of intravenous (IV) patient doses involves the process of dissolving a drug for administration to a patient and/or diluting or mixing the drug with an infusion solution used as a vehicle for the purpose of administration.
Bevacizumab, infliximab, trastuzumab and rituximab are widely used mAbs. The dose is calculated based on the indication and patient characteristics, such as body weight or surface area. From a compounding perspective, the consequence of these individual patient doses is the introduction of remainders. As an example, a patient dose of 450mg infliximab (= 45ml of 10mg/ml) will introduce a remainder of 50mg (= 5ml) because the standard vial size is 100mg. For each mAb, the amount of remainder after compounding a patient dose depends on the sizes of vials that are available.
Trastuzumab and infliximab are available in only one vial size (in Europe), whereas bevacizumab and rituximab come in two sizes. Regardless of the currently available vials, valuable remainders are wasted daily.
In practice, hospitals have made several efforts to reduce spillage.
Methods used most frequently are: (i) clustering of patients on one or more days; and (ii), dose banding. Clustering of patients for treatment on one day instead of every (working) day results in wasting one remainder a week instead of five. This strategy, however, compromises patient service, given that patients can only be administered their drug on the designated day(s) of the week. Furthermore, the process of manual compounding using remainders from a previous patient the same day, introduces a risk of mix-ups, because there is no compliance to the principle of line clearance within the laminar air flow (LAF) cabinet.
Nevertheless, the financial consequences of not clustering patients might outweigh the risk of a mix-up. However, there are also hospitals that do comply completely with the principal of line clearance, which means that they do not use remainders of one patient for the next one on the same day when other drugs are compounded in the same LAF cabinet in between. Dose banding means that a dose is rounded up or down to prevent introduction of remainders. Dose banding can, however, lead to >10% deviation from the dose that should have been administered according to the registered dose.
Spillage of remainders of mAbs represents a significant cost. We compound approximately 3000 mAb patient doses per year in our hospital, of which >50% include trastuzumab, infliximab and bevacizumab. Theoretically, this means that at the end of each working day, 50% of the smallest vial size available of each mAb is wasted on average. In our hospital, we monitored the waste of remainders for a few weeks and found wastage equivalent to costs of approximately €2800 per week (trastuzumab, infliximab and bevacizumab only).
Compounding robots
Traditionally, manual compounding of drugs for IV administration is an activity carried out by pharmacy technicians and nurses. Recently, partially and fully automated robots have come to the market as safe, efficient, and cost-effective alternatives for manual compounding of drugs for IV administration. In addition, these innovations have introduced the possibility to re-use remainders by means of a proper track-and-trace system. The i.v.STATION® (Health Robotics SRL, Bolzano, Italy) is a compounding robot (GMP grade A) that includes this system and allows us to: store and reuse partially used vials during a full working day without a risk of mix-ups and unload a partially used vial at the end of the day.
All unloaded vials are labelled with a barcode that holds the information about when the drug was reconstituted/first punctured, how much was used, how much is left and the expiration date for the vial. The robot will refuse the vial if the expiration date has passed. The time to expiration of a mAb depends on microbiological and chemical stability. Extensive microbial validation confirmed sterility for at least seven days after first puncture of a vial in the robot.
The chemical stability of reconstituted infliximab and trastuzumab has also been shown to be at least seven days, which allows unloading and reloading within a week.(2,3) Stable liquid formulations are chemically stable by definition.
So far, the performance of automated solutions for compounding is valued predominantly for its timesaving capacity. Rapid procedures, however, might harm delicate substances such as mAbs, and may therefore be inappropriate. Specific instructions for compounding are provided in many of the Summaries of Product Characteristics for mAbs. They state that gentle swirling is required to aid reconstitution (in case of lyophilised powder) and drawing into a syringe should be performed slowly after settling for the instructed amount of time.(4,5) Foaming and formation of protein aggregates may occur if the reconstitution procedure is not performed accordingly.(6) Aggregates have been associated with immune reactions.(7) Moreover, foaming might hamper drawing correct volumes into a syringe.
Our study
We performed a study to assess whether a fully automated robotic procedure can achieve similar quality compared with manual compounding of mAbs.(8) Three commonly used MAbs were studied: bevacizumab, infliximab and trastuzumab.
In brief, we first developed a swirling protocol using a small stand-alone robotic arm able to make exactly the same movements in 3D space as the robotic arm installed in the i.v.STATION®. The swirling protocol was developed to resemble the manual swirling procedure instructed most in the SPC. In addition, the protocol was tested to completely dissolve the lyophilised mAbs (infliximab and trastuzumab), assessed by visual inspection. Second, the lyophilised MAbs were reconstituted using three different procedures: (i) manual reconstitution according to the SPC; (ii) robotic reconstitution using the swirling protocol; and (iii), a vigorous swirling method (worst case scenario). The aggregation state for all samples was assessed using five different techniques able to detect small differences in aggregation state of protein formulations.
The samples that were reconstituted using procedure III showed clear changes in aggregation state for both infliximab and trastuzumab. No aggregation was detected by any of the analytical methods for the vials reconstituted by the manual solubilisation procedures according to the SPC and the automated method by the robotic system. These results confirmed that the i.v.STATION® robot can compound mAbs in the correct way without introducing any risk of protein aggregation.
Conclusions
Automated methods may provide a solution for mAbs that have not been marketed because of formulation problems. When a mAb enters the market, many challenges have been overcome to obtain a robust formulation with low aggregation propensity.(9) Little attention is paid to aggregation properties during the discovery phase for a mAb. Once the sequence of a mAb is fixed, it enters the development phase. mAbs often have high aggregation properties, thereby introducing a major challenge in the design of a stable formulation, and at high costs. Efforts are made to predict aggregation properties and to direct selection of mAbs during early development by means of a developability index(10) highlighting the significance of aggregation problems often faced in the formulation stage.
Factors that can influence aggregation state in the formulation stage include additives (design), pH, shear stress, interaction with package material such as glass and rubber stopper (manufacturing), temperature, shelf life (storage), and inappropriate compounding at the site of administration. We have shown that a compounding robot can be programmed to perform complex procedures (for example, the right angle to add the water onto the lyophilised cake, the right drawing and swirling speed, and settling time) for each individual mAb product and provide a solution for mAbs requiring very specific handling during compounding.
Key points
- Many monoclonal mAbs are dosed based on patient characteristics. Consequently, valuable remainders are often wasted.
- Partially and fully automated robots have come to the market as an alternative for manual compounding. Safely reusing valuable remainders is possible by means of a proper track-and-trace system, thereby reducing spillage.
- Specific instructions on for compounding are provided in many of the Summaries of Product Characteristics of mAbs.
- Timesaving automated procedures may harm delicate substances such as mAbs. We showed that a fully automated robot is capable of compounding high-quality intravenous mAb patient doses.
- Aggregation properties of mAbs may hamper market access due to formulation problems. Automated methods for compounding can provide control of parameters that are known to introduce aggregation when compounding (for example, shear stress, resting time, contact with rubber stopper).
References
- Nelson AL, Dhimolea E, Reichert JM. Development trends for human monoclonal antibody therapeutics. Nat Rev Drug Discov 2010;9:767–74.
- Food and Drug Administration. Herceptin (trastuzumab) lLabel. www.accessdata.fda.gov/drugsatfda_docs/label/2010/103792s5250lbl.pdf (accessed 8 August 2013).
- Beer PM et al. Infliximab stability after reconstitution, dilution, and storage under refrigeration. Retina 2010 Jan;30(1):81–4.
- Remicade. Summary of Product Characteristics. www.ema.europa.eu/docs/en_GB/document_library/EPAR_-_Product_Information/human/000240/WC500050888.pdf (accessed 8 August 2013).
- Herceptin Summary of Product Characteristics. www.ema.europa.eu/docs/en_GB/document_library/EPAR_-_Product_Information/human/000278/ WC500074922.pdf (accessed 8 August 2013).
- Demeule B et al. New methods allowing the detection of protein aggregates: A case study on trastuzuMAb. MAbs 2009;1:142–50.
- Hermeling S et al. Structure-immunogenicity relationships of therapeutic proteins. Pharm Res 2004;21:897–903.
- Peters BJ et al. Validation of an automated method for compounding monoclonal antibody patient doses: case studies of Avastin (bevacizumab), Remicade (infliximab) and Herceptin (trastuzumab). MAbs 2013;5(1):162–70.
- Lowe D et al. Aggregation, stability, and formulation of human antibody therapeutics. Adv Protein Chem Struct Biol 2011;84:41–61.
- Lauer TM. Developability index: a rapid in silico tool for the screening of antibody aggregation propensity. J Pharm Sci 2012;101(1):102–15.
