Objectives Aggregation is one of the key critical points limiting the stability of monoclonal antibodies in solution. The present study aimed to investigate the in-use stability of a residual monoclonal antibody solution after withdrawal of most of the filling volume of PF-06439535 (bevacizumab biosimilar), addressing the physical and chemical stability with respect to aggregation and fragmentation.
Methods The stability of residual PF-06439535 solution (25 mg/mL) after withdrawal of 80% (12.8 mL) filling volume with a 20G needle was monitored over a light-protected storage period of 8 days at 2–8°C and 25°C with measurement time points at D0 (start of storage), D2, D4, and D8 (2, 4, and 8 days of storage after start, respectively). Unopened vials stored under the same conditions served as control. For this purpose, the analytical results from size exclusion chromatography, dynamic light scattering, and micro-flow imaging obtained after the individual measurement time points up to 8 days were compared with those obtained at D0 and with those obtained for unopened vials stored under the same conditions.
Results No aggregation or ongoing fragmentation due to partial withdrawal of filling volume could be observed in the residual PF-06439535 solution. Moreover, no changes in the particle size distribution at D8 compared with the D0 values were identified upon storage at either 2–8°C or 25°C (both opened and unopened vials). The total concentration of particles ≥10 µm of all samples was <100 particles/mL. In addition, no variations in the pH values or in the visual appearance were detected over the whole study period in all samples at all storage conditions.
Conclusions Consequently, residual PF-06439535 solution (25 mg/mL) in opened vials may be regarded as stable when stored light-protected over a period of 8 days in the refrigerator (2–8°C) or at 25°C.
- CLINICAL MEDICINE
- MEDICAL ONCOLOGY
- PHARMACEUTICAL PREPARATIONS
- Drug Monitoring
This is an open access article distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited, an indication of whether changes were made, and the use is non-commercial. See: http://creativecommons.org/licenses/by-nc/4.0/.
Statistics from Altmetric.com
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.
WHAT IS ALREADY KNOWN ON THIS TOPIC
For Avastin, the first bevacizumab monoclonal antibody registered in the EU, degradation of bevacizumab was <13% in pierced and sampled vials when stored at 4°C for 6 months.
PF-06439535 is a bevacizumab biosimilar; there is a great demand in clinical practice for stability data of the remaining bevacizumab concentrate solutions in sampled large volume vials of PF-06439535.
WHAT THIS STUDY ADDS
No changes in the physical and chemical stability of residual PF-06439535 concentrate (25 mg/mL) could be observed after partial volume (12.8 mL of 16 mL vial, 80% of the filling volume) was withdrawn and stored protected from light in a refrigerator at 2–8°C or in an incubator at 25°C for 8 days.
HOW THIS STUDY MIGHT AFFECT RESEARCH, PRACTICE OR POLICY
There is a strong interest in clinical practice in the stability of residual PF-06439535 concentrate in opened vials. Based on the generated extended stability data obtained in the present study, the stability of PF-06439535 should be verified in more extensive studies.
Monoclonal antibodies (mAbs) are among the most important classes of therapeutic proteins, in particular for oncology and rheumatoid and autoimmune diseases.1 One of the most common mAbs used to treat cancer is the anti-vascular endothelial growth factor (anti-VEGF) antibody bevacizumab because of its anti-angiogenic activity.2–5 Bevacizumab is additionally often used off-label as cost-effective anti-VEGF therapy in ophthalmology.6
The first bevacizumab mAb registered in the EU was Avastin in January 2005.7 Avastin has been approved to treat a range of cancers in adult patients.7 8 After loss of exclusivity of the originator bevacizumab, several biosimilars were launched to improve patient access to bevacizumab by increasing treatment options and reducing treatment costs. To be approved, biosimilars must demonstrate similarity to reference products in analytical, non-clinical, and clinical pharmacology as well as clinical efficacy/safety studies; and they must have no clinically meaningful differences in quality, safety, and efficacy compared with their reference product.9 10 PF-06439535 (Zirabev; Pfizer) is a bevacizumab biosimilar. Similarity was established through a rigorous comparison of PF-06439535 and the reference product.11–14 PF-06439535 was approved by the European Medicines Agency in February 201915 and launched in Germany in June 2020.16 PF-06439535 is commercially supplied in 4 mL and 16 mL vials containing 100 mg and 400 mg bevacizumab, respectively.16 These are concentrates for the preparation of diluted infusion solutions. Since the clinical doses are dependent on the indication and patient’s body weight,16 the whole volume is often not drawn and residual volumes are left in the vial (especially the 16-mL vials). In the summary of product characteristics it is stated that bevacizumab vials are intended for single-use only and any residual volumes have to be discarded immediately.16 Nevertheless, there is a great demand in clinical practice to obtain stability data on the residual volume of bevacizumab concentrate solutions in opened large volume vials of PF-06439535.
Many reports exist for the originator on bevacizumab stability and sterility when repackaged in syringes for intravitreal application.17–20 The stability of residual bevacizumab concentrate volumes in fractionated vials has also been investigated,18 revealing <10% degradation and preservation of the anti-VGEF activity, as well as sterility for up to 6 months, when the remaining content of multiple-dose vials of the originator product (Avastin) was stored at 4°C and aseptic conditions were followed during sampling. Another study reported a 9.6% and 12.7% degradation of bevacizumab in pierced and sampled vials of the originator product compared with unpierced unopened vials at 3 and 6 months, respectively.17
Due to the complexity of mAbs in terms of their structure and proprietary nature of their manufacturing process, some levels of heterogeneity usually exist not only between batches of the originator biologic product but also between biosimilars and the originator. Therefore, the stability data already collected for residual bevacizumab concentrates in fractionated vials of the originator product cannot be extrapolated to PF-06439535. Based on that background, we looked at the physical and chemical stability of bevacizumab PF-06439535 solution (25 mg/mL) with regard to fragmentation and aggregation over a period of 8 days in vials from which 80% (12.8 mL) of the initial volume (16 mL) was removed. These vials were stored protected from light at 4°C and 25°C, respectively, and subjected to stability analysis by applying size exclusion chromatography (SEC), dynamic light scattering (DLS), and micro-flow imaging (MFI), along with visual inspection of the appearance and measurement of pH.
Chemicals, reagents, and equipment
Commercially available Zirabev 16 mL concentrate vials for infusion (PF-06439535) containing 400 mg bevacizumab-bvzr (Lot DK 1277) with expiry date 03/2023 were provided by Pfizer Pharma GmbH (Berlin, Germany). PF-06439535 is formulated at 25 mg/mL and excipients include succinic acid, sucrose, edetate disodium dihydrate (EDTA), polysorbate 80, sodium hydroxide to titrate solution to pH 5.5, and water for injection.16 21 The protein standard mix consisting of ribonuclease A type I-A from bovine pancreas (1 mg/mL, 13.7 kDa), albumin chicken egg grade VI (1 mg/mL, 44.3 kDa), and thyroglobulin bovine (0.5 mg/mL, 670 kDa) used for calibration was bought from Sigma Aldrich (Darmstadt, Germany). Thermo Scientific COUNT-CAL count standard 10 µm, Thermo Scientific Nanosphere Particle size standard 20 nm, and Thermo Scientific Nanosphere Particle size standard 50 nm were purchased from Fisher Scientific (Schwerte, Germany).
Potassium dihydrogen phosphate, sodium chloride, and sodium hydroxide were purchased from Merck (Darmstadt, Germany). Purified water (as defined by the United States Pharmacopeia) was purified by using a Milli-Q purification system (Merck Millipore Corporation, Bedford, MA, USA).
SEC was carried out on an Agilent 1200 liquid chromatography system equipped with a binary pump SL of the Agilent 1200 series, an autosampler (Agilent 1200 series G1367C with coupled thermostat G1330B), and an ultraviolet detector (Agilent 1200 series G1315C). MFI analysis was conducted on an MFI-5200 particle analyser system (ProteinSimple, Santa Clara, CA, USA) equipped with a silane-coated high-resolution 100-µm flow cell. DLS was measured on a Zetasizer Nano ZS system (Malvern Instruments, Worcestershire, UK) using single-use 220–1600 nm UVette plastic cuvettes from Eppendorf (Hamburg, Germany).
A total of 84 PF-06439535 16 mL vials were received and stored for approximately 1 week at 2–8°C until study initiation.
At the start of the study, approximately 80% of the filling volume (12.8 mL) was withdrawn from the vials using a 20G needle connected to a 10-mL syringe. The pierced vials were not covered by parafilm but were stored protected from light upright at 2–8°C in a temperature-controlled refrigerator. An equivalent set of samples was kept in a temperature-controlled incubator at 25±2°C with 60±5% relative humidity. Analysis was carried out at the starting point (D0), and at the second (D2), fourth (D4), and eighth (D8) day of storage. Unopened (ie, not punctured) control vials were stored in the same refrigerator and incubator to be tested at D8.
At each time point the respective number of vials were removed from the storage location and analysed on the same day. The vials stored in the refrigerator were equilibrated to room temperature before analysis. Any necessary dilutions were carried out immediately before analysis. The analytical investigations are shown in table 1.
The samples were visually inspected for the presence of visible particles under a white light in front of a matte black panel and a non-glare white panel.
pH measurements were made using a pH metre (SevenMulti pH/LF S47-KS with InLab Routine Pro electrode; Mettler Toledo, Gießen, Germany) calibrated under European Pharmacopoeia conditions. For each storage period, the pH measurements were performed in triplicate for both storage conditions.
Any formation of aggregates (high molecular weight impurities) or fragments (low molecular weight impurities) was detected by SEC using a Thermo Scientific MabPac SEC1 300 Å, 4.0×300 mm, 5 µm column, and Thermo Scientific MabPac SEC1 300 Å, 4.0×50 mm, 5 µm precolumn (Fisher Scientific, Schwerte, Germany). The samples were diluted with mobile phase to a concentration of 0.5 mg/mL for the determination of molecular weight and 5.0 mg/mL for the determination of side peaks via peak area percent. Three solutions were prepared from three vials, respectively, and each solution was measured twice. The mobile phase consisted of 100 mM potassium dihydrogen phosphate, containing 200 mM sodium chloride adjusted to pH 6.8 with sodium hydroxide. Chromatographic separation was carried out isocratically at a flow rate of 0.3 mL/min, applying an injection volume of 1 µL and adjusting the column temperature to 30°C and the autosampler temperature to 15°C. The isocratic elution profile obtained over a run time of 30 min was monitored using ultraviolet detection at 214 nm. The SEC column was calibrated by measuring the elution volume of molecular weight markers commercially available in a protein standard mix solution, the composition of which is mentioned above. Log molecular weight values of the standards were plotted against the elution volume to determine the equivalent molecular weight of the sample.
The method was validated by verifying linearity, repeatability, and accuracy. For verifying linearity, five different dilutions of the bevacizumab PF-06439535 sample (0.5, 1, 2, 4, and 5 mg/mL) were analysed in duplicate, respectively, yielding R2=0.98832. The repeatability was verified by six-fold measurement of the PF-06439535 sample at both concentrations of 0.5 mg/mL and 5 mg/mL, yielding variation coefficients corresponding to 0.17% and 0.09%, respectively. For verifying accuracy, the molecular weight of five different dilutions of the PF-06439535 sample (0.5, 1, 2, 4, and 5 mg/mL) were analysed in duplicate, yielding molecular weights not deviating by more than 10% from the declared molecular weight. Based on the signal to noise ratio, the limit of quantitation was determined to correspond to 0.025%. A detailed overview of the individual values is presented in the online supplemental tables S4–S6 and in the online supplemental figure S1.
For DLS analysis, the samples were diluted with purified water to a final protein concentration of 10 mg/mL; 250 µL of the diluted sample were pipetted into the cuvette and each sample was equilibrated to 20°C for 120 s before measurement.
For stability testing, three vials at each time point and storage condition were analysed in triplicate by applying the automatic measurement mode at 20°C at an angle of 90°. Instrument parameters, such as attenuator settings, were optimised by the software to obtain the recommended count rates.
Malvern Zetasizer Software version 7.03 was used to fit the autocorrelation function by using the default settings for protein solutions and water (refractive index=1.333; viscosity=0.890 mPa*s) to calculate the Z-average diameter, polydispersity index (PdI), and particle size distribution by intensity. The average of the triplicate analyses was reported as analytical results.
At each day of measurement, a system suitability test was performed with two sizing standards (Nanosphere series 3000, sizing standards 20 nm and 50 nm).
Samples were analysed undiluted in MFI. The background illumination was performed using purified water. Samples of 1.0 mL were loaded and a pre-run volume of 0.25 mL was followed by a sample run of 0.58 mL at a flow rate of 0.17 mL/min. MFI View System Software version 2-R2-188.8.131.525 was used to perform the measurements, and MFI View Analysis Suite (MVAS) version 184.108.40.2067 was used to analyse the samples for the presence of subvisible particles in a size range between 2–100 µm.
For the discrimination of silicone oil droplet-like particles (SP) and protein particles (non-silicone oil particles (NSP)) in the size range starting from 5 µm, the ‘find similar’ operation in the MVAS software was used. SP were selected manually as a basis for the automatic search function by the software. The particles were then categorised into three groups: the total number of particles, SP only, and NSP only, including the respective images.
Before sample measurement, a particle count control measurement was performed using a polystyrene counting standard of a defined particle size and particle concentration.
Student’s t-test at the common significance level α=0.05 was applied to compare the individual values obtained by SEC, DLS, and MFI for particles ≥2 µm for the opened vials at the different time points at different storage conditions, and for the unopened vials at different storage conditions with the D0 values, respectively.
The visual assessment of the vials stored in the refrigerator or at 25°C revealed clear, colourless solutions, free of visible particles.
No significant changes in pH values were observed. The average pH values of opened vials were 5.49 at D0 and D8 when stored at 2–8°C and between 5.49 (D0) and 5.47 (D8) when stored at 25°C, being comparable to the D8 pH values of the unopened vials when stored at 2–8°C (5.48) and at 25°C (5.47).
Using the MabPac SEC1 300 Å column, the molecular mass of PF-06439535 could be determined accurately indicating the proper implementation of the method following the calibration with known molecular weight standards.
As can be depicted from the SEC chromatograms presented in figure 1A and B, the chromatographic profiles obtained for PF-06439535 after 8 days of storage at 2–8°C and 25°C were comparable to those at D0. Moreover, there was no difference between samples from opened or unopened vials.
Also, no additional side peaks were detected when comparing the SEC chromatograms of PF-06439535 diluted to a concentration of 5 mg/mL after storage for 8 days at 2–8°C and 25°C, respectively, to D0. The visible side peaks 1, 2, and 3 were already present at the start point. The same applies for the unopened vials (see figure 1C). Only peak 3a could not be seen at D0. It is assumed that it probably co-eluted with peak 3 at the start point and eluted at a later retention time in the following time points. For peak area evaluation, peaks 3 and 3a were calculated as sum.
At the start, side peak 1 was present in a size of about 0.06 area percent. During stability analysis, it slightly decreased to values below 0.04% in the unopened control vials as well as in the opened stability vials stored in the refrigerator or at room temperature. This side peak eluted earlier than PF-06439535, which means that this compound has a higher molecular weight.
Side peak 2 also eluted earlier than PF-06439535 in a quite constant size of 2.8–3.0% at all conditions and time points tested.
Side peak 3 (with 3a) eluted later than PF-06439535, indicating a lower apparent molecular weight. As peak 3a was also detected in the unopened control vials, it is assumed that the appearance of this peak may be attributed to a change in the binding performance of the SEC column rather than to an ongoing degradation effect after piercing the vials and withdrawing 80% of the filling volume. This assumption is further supported by the fact that the peak area was quite constant from D2 until D8 under all conditions tested. An overview of the average individual area percentages obtained for the individual side peaks is presented in table 2, and a detailed overview of the individual values is presented in the online supplemental material tables S1–3.
Statistical analysis revealed significant differences for all values listed in table 2 compared with the respective D0 values for all side peaks except for side peak 1 at D0/2–8°C. Nevertheless, in no case did the sum of the three side peaks exceed an area percentage of 4.541% compared with the PF-06439535 peak. This indicates that the specified range of ±10% for PF-06439535 is complied with under all testing conditions. Consequently, PF-06439535 is considered stable for the investigated storage time and condition, even if statistical differences are detected.
The size distributions by volume for storage at 2–8°C and 25°C (figure 2A and B) suggest that all analysed samples were characterised by one dominating main species (≥98% main peak) at about 11–12 nm, which is the expected size for an mAb.22 23 No differences were observed for up to 8 days of storage at 2–8°C and 25°C. The additional unopened vials analysed at D8 exhibited a similar profile as compared with the opened vials.
A summary of Z-average diameters and PdI values calculated from triplicate DLS measurements for all tested samples at all stability time points is presented in table 3. Statistical analysis revealed no significant changes in the Z-average values when comparing the individual values for the opened vials at the different time points at different storage conditions, and for the unopened vials stored at 25°C for 8 days with the D0 values. Only the unopened vial stored at 2–8°C revealed a significant difference to the D0 value, which might be attributed to inherent different Z-average diameters in both unopened vials.
No substantial changes or trends in sample polydispersity or Z-average diameter were determined in the PF-06439535 samples stored at 2–8°C or 25°C for up to 8 days. At D8, the vials with a residual filling volume of approximately 20% showed similar results as compared with unopened control vials.
MFI measurements were performed to evaluate the particle concentrations in the larger subvisible size range of approximately 2–100 µm. In general, low levels of subvisible particles were found in the samples. No clear trend of changing particle concentrations upon storage at both temperatures or between opened and unopened vials was identified (table 4).
Statistical analysis was solely carried out on particles ≥2 µm as they best represent any possible trend. No significant changes in the total particle concentration could be detected when comparing the individual values for the opened vials at the different time points at different storage conditions, and for the unopened vials stored at different storage conditions with the D0 values. The differentiation between SP and NSP is feasible for particles >5 µm based on their characteristic appearance.24 Overall, the majority of particles were NSP for all sampling time points.
Bevacizumab is labelled as stable for 3 years in the source vials when stored unopened at 2–8°C, protected from light and not frozen.16 Furthermore, in-use stability data are provided in the summary of product characteristics of PF-06439535 (Zirabev) for the diluted concentrate in 0.9% saline solution, but not for the concentrate remaining in the opened vial.16 In protein solutions like bevacizumab, concentrate aggregation can only occur with temperature changes, protein concentration changes, or mechanical stress.25 In pharmaceutical practice, storage temperature and protein concentration are generally well controlled, representing no risk factors. However, the impact of mechanical stress from partial volume withdrawal on the quality of the remaining solution is not known. Since the collision of native and/or denatured proteins can be considered the major step in protein aggregation, it cannot be excluded that withdrawing a certain volume from the vial may affect the stability of the residual solution.26
The summary of product characteristics of PF-06439535 states that the product is for single-use only and any unused medicinal product should be disposed of since it contains no preservatives.16 Nevertheless, there is a strong interest in clinical practice in the stability of the residual bevacizumab concentrate solutions in sampled large volume vials. Therefore, the present study aimed to evaluate the physical and chemical stability of fractionated PF-06439535 concentrate samples following the withdrawal of 80% (12.8 mL) of the prefilled volume stored under various storage conditions over a period of 8 days, applying different complementary stability-indicating analytical techniques.
SEC-high-performance liquid chromatography (HPLC) was used to detect a possible formation of aggregates and fragments, as this technique can separate the mAb monomer from the polymeric form (dimers or more) and fragments. DLS allowed for estimation of the presence and size distribution of any formed protein aggregates in the nanometre size range. MFI was applied to quantify subvisible particles >1 µm and distinguish between different particle types such as protein aggregates, SP, and air bubbles in the size range ≥5 µm, thereby providing insight into the nature of the observed particles.
A limitation of this study is that microbiological tests were not conducted in order to exclude any microbiological contamination of the residual solution after withdrawal of 80% of the volume. Another limitation is the concentration on size exclusion chromatography as the only method for evaluating the chemical stability of bevacizumab. Of course, ion exchange chromatography could have been used to monitor the stability of bevacizumab PF-06439535. Since, however, the aim of the study was to establish a preliminary idea of the effect of volume withdrawal on the stability of PF-06439535, and since it is clearly recommended in the summary of product characteristics16 to dispose of any unused residual volume, we are of the opinion that SEC is sufficient for the purpose of the present study. Therefore, the study was designed so that stability assessment of the fractionated vial was deduced from the comparison of the analytical results obtained at D0 with those obtained at different time points and at the end of the 8 day storage period. In addition, unopened vials serving as control and stored under the same conditions as the fractionated vials for the same period of time were also comparatively evaluated.
Based on the results obtained in the present study, no aggregation or ongoing fragmentation due to partial withdrawal of filling volume could be observed in SEC-HPLC. Moreover, no changes in the DLS particle size distribution at D8 compared with the D0 values were identified upon storage at either 2–8°C or 25°C, neither in the pierced nor in the unopened vials. The total concentration of particles ≥10 µm of all samples was <100 particles/mL; the other size ranges showed comparable results. In addition, no significant variations in the pH values or in the visual appearance were detected over the whole study period in all samples at all storage conditions. Based on the above, no changes in the physical and chemical stability could be observed for PF-06439535 formulation (25 mg/mL) over a period of 8 days when stored protected from light at 2–8°C or 25°C after partial withdrawal of filling volume.
Based on the results of several complementary stability-indicating analytical techniques including SEC, DLS, and MFI, no changes in the physical and chemical stability of residual PF-06439535 (Zirabev) concentrate (25 mg/mL) could be observed after partial volume withdrawal of 80% of the filling volume and stored protected from light in the refrigerator at 2–8°C or at 25°C for 8 days.
Patient consent for publication
We thank Angel H Bair, Martin J Summers and Raj Patel from Pfizer for reviewing the manuscript.
EAHP Statement 5: Patient Safety and Quality Assurance.
Correction notice This paper has been amended since it was first published. Affiliation 3 has been corrected.
Contributors MW, JW, AH, and FG designed and carried out the experiments. MA-T wrote the draft text. All authors participated in the design of the study and manuscript review process. All authors approved the final version of the manuscript for submission.
Funding This study was sponsored by Pfizer. Editorial support was provided by Shuang Li, PhD, of Engage Scientific Solutions and was funded by Pfizer. Open access publication of this article was funded by Pfizer.
Competing interests MA-T and MW are employees of the Central Laboratory of German Pharmacists. JW was an employee of the Central Laboratory of German Pharmacists when the study was carried out. AH is an employee of and has stock and/or other ownership interests in Coriolis Pharma Research GmbH, Germany. FG was an employee of Coriolis Pharma Research GmbH, Germany. AMS is an employee of and has stock and/or other ownership interests in Pfizer. JHS was an employee of Pfizer.
Provenance and peer review Not commissioned; externally peer reviewed.
Supplemental material This content has been supplied by the author(s). It has not been vetted by BMJ Publishing Group Limited (BMJ) and may not have been peer-reviewed. Any opinions or recommendations discussed are solely those of the author(s) and are not endorsed by BMJ. BMJ disclaims all liability and responsibility arising from any reliance placed on the content. Where the content includes any translated material, BMJ does not warrant the accuracy and reliability of the translations (including but not limited to local regulations, clinical guidelines, terminology, drug names and drug dosages), and is not responsible for any error and/or omissions arising from translation and adaptation or otherwise.