Contamination risk? Examining the potential risk of single-use plastic bioprocess bags

With a trend moving towards more use of single-use bioprocess bags to transport biopharmaceutical solutions Dujuan Lu, PhD (technical client manager/global lead, SGS Life Science Services) and Kate Comstock (senior marketing specialist, Thermo Fisher Scientific) assess whether or not contaminants leach out from these bags, potentially creating a risk to patient safety.

Single-use bioprocess bags are increasingly being used for the preparation, storage and transport of biopharmaceutical solutions, intermediates and final bulk products. These bags are made from plastic materials, and there is the risk that contaminants, additives and degradants may leach out into the cell culture media and remain inside final drug products.

It is therefore an important part of the safety risk assessment process that any such leachable is both identified and quantified. Leachable contaminants pose a risk to the cell culture process itself and can be toxic to cells used in biopharmaceutical production. Moreover, leachable contaminants may directly pose risks to patients, either through direct toxicity or because they may alter the efficacy of the medicinal products.

Extractable study

SGS, in collaboration with Thermo Fisher Scientific, carried out an extractable study on plastic single-use bioprocess bags using different solvent systems, including acidified water, alkaline water, phosphate buffered saline (PBS), and mixtures of organic and aqueous solvents. The aim was to bracket and mimic the pH values, ionic strength and hydrophobicity of common process fluid solutions, then multiple analytical techniques were used to give a comprehensive profile of the extractables. These included headspace analyses using GC-FID & GC–MS for volatile organic compounds, liquid injection GC-FID & GC–MS analysis for semi-volatile organic compounds, and LC-UV & LC–MS analysis for non-volatile organics.

The analysis was carried out on commercially available plastic bioprocess bags. Extractions were performed on the bags using water at both pH3 and pH9, PBS, 1:1 mixtures of both isopropyl alcohol (IPA) and ethanol with water. Bags filled with the five different solutions were placed in a shaker at 50 °C for seven days, and the extract solutions then analysed using the different techniques.

Analyses

Headspace GC–MS analysis was used to identify and quantify volatile organic compounds using an Agilent 6890N machine. Cyclohexanone was found in all of the extracts; this was also found using liquid-injection GC–MS analysis. The overlap in extractable profile by these two techniques reiterates the completeness of the volatile and semi-volatile organic compound profile determined via this method.

IPA was detected in all of the aqueous extracts via headspace analysis. The presumption was that vapour from the IPA/water extract in the bags could migrate into the aqueous solvents within other bags inside the same shaker. Further studies were carried out, with the bags containing aqueous solvents and those with organic solvents being separated within the shakers, and this time, no IPA was detected in the aqueous extracts. Therefore, in future it will be important to keep the extracts separated to prevent this type of contamination. Covering the bags with aluminium foil without separating the bags also greatly decreased the IPA levels in the aqueous extracts.

GC–MS analysis was used to identify and quantify semi-volatile organic compounds, using a Thermo Scientific ISQ LT single quadrupole GC–MS system with National Institute of Standards and Technology (NIST) library, and a Thermo Scientific Q Exactive high resolution accurate mass (HRAM) GC–MS/MS system to aid identification. The aqueous extracts are usually solvent-exchanged with dichloromethane (DCM) before injection, but emulsification can be problematic with liquid–liquid extraction if there is a high amount of alcohol in the samples. The effect of solvent exchange was therefore studied by injecting the ethanol/water and IPA/water extracts both with and without DCM extraction, as shown in Figures 1–4. Better peak shape and higher recovery were observed in the samples with DCM extraction, regardless of emulsion and phase separation issues.

The highest concentration of extractables was found in the IPA/water extract, particularly hydrophobic compounds. The major peak in every extraction was bis-(2-ethylhexyl) phthalate, or DEHP, which is the primary plasticiser used in these plastic bags. The majority of the rest of the extractables were phthalates, degradation products from DEHP, and lubricants.

LC–MS analysis was used to identify and quantify non-volatile organic compounds, using a Thermo Scientific Q Exactive high resolution accurate mass (HRAM) LC–MS/MS system. Identification of unknown compounds in the LC–MS analysis for non-volatiles was more challenging as there is currently a dearth of standardised spectral libraries and LC–MS spectra vary from vendor to vendor significantly. Instead, a high-resolution accurate mass (HRAM) full scan and MS/MS data acquisition with polarity switching was used to aid structure elucidation. This technique ensured the detection of structurally diverse compounds, provided a comprehensive extractable profile, and increased the analysis throughput.

As can be seen in Figure 5, 20 extractables are evident in the IPA/water extract, significantly more than was seen in any of the other four solvents, both in terms of number of peaks and concentration. To identify the compounds, the HRAM data was processed using Thermo Scientific Compound Discoverer small molecule structure analysis software. This gives structure predictions and then performs an online library search against ChemSpider (RSC) and mzCloud (HighChem), as well as a local extractables and leachables compound database search. Again, the major peak observed from all extracts was DEHP.

Higher concentrations of hydrophobic compounds were detected in the IPA/water extract. There were two extractables in the IPA/water extract with the same molecule weight and chemical composition, but which eluted with different retention times, indicating that they were structural isomers. An MS/MS analysis indicated a different fragmentation pattern. Running both full scan and MS/MS analysis will be required to get the full picture.

Importance of comprehensive investigations

The study highlighted the importance of carrying out careful, comprehensive investigations into how plastic bioprocess bags being used for biopharmaceutical products behave when in contact with liquids that might extract compounds from them. Only by understanding what might be extracted, and the quantities that are likely to be extracted under what conditions, can the risks that extractables may pose to patients be assessed and quantified.