Chasing Traces

Supply chain integrity and reliability is critical for the biopharma industry, perhaps no more so than in the manufacture of cell culture media. A given medium may comprise of 50 to 80 different raw materials, and low levels of impurities in each component can have a cumulative impact on the final medium composition. Impurities can affect multiple pathways of the cells that are grown in cell culture medium, thus contributing to the variability of proteins harvested from those cells. Some trace metals impact certain glycosyltransferases and can alter the protein glycosylation profile. In particular, concentrations of trace elements like copper, manganese, zinc, and selenium, are absolutely key because they have a direct impact on protein quality. Other trace metals are critical nutrient sources in their own right – iron in particular is essential for cell growth. Whether the trace metal is intentional in the media formulation or an impurity, trace components have different effects and “ideal” concentrations may vary according to the process in question. To avoid product quality issues, it is vital that biopharma and biosimilars companies understand the effect of elemental metals on a given bioprocess, and quantify the impurities present in their processes.

Regulatory goals

Recognition of the critical impact of trace metals is reflected in evolving regulatory guidelines. The pharma industry is currently adjusting to new guidelines, such as the FDA’s ICH-Q3D “Elemental Impurities” document, concerning acceptable impurity levels in drug products. Broadly, regulators now favor replacement of traditional analytical chemistry methods with more sensitive techniques for trace metal quantification, such as inductively-coupled plasma mass spectrometry (ICP-MS). Industry must familiarize itself with these methods – and media suppliers must adapt to this trend. So how is MilliporeSigma positioned in this environment?

We now have a state-of-the-art trace metal analysis facility – the result of significant investment and a real development journey. When we first started fully characterizing raw materials, we didn’t initially think about trace metals; however, on closer inspection, we saw surprising variability in elemental impurities and realized that we needed to expand our dataset to make sure we understood our raw materials relative to trace metal impurities. As we collected more data, it became increasingly clear that the issue needed serious attention. Unfortunately, the external analytical laboratory that we used at the time wasn’t providing us with the data we needed – they worked at a parts per million sensitivity when we needed parts per billion. Eventually, we made the decision to develop an in-house, dedicated facility for the analysis of elemental impurities. We renovated our existing space, procured a high quality ICP-MS instrument, hired some personnel, and started generating our own data. We chose ICP-MS as the workhorse analysis method because it is about as exact and quantitative as you can get, but we do sometimes also use ICP-AES (atomic emission spectroscopy) or ICP-OES (optical emission spectroscopy), depending on the quantity of the element we are studying. In general, however, we rely heavily on ICP-MS.

In short, we quickly went from not even knowing that we should look at trace metals to a purpose-built, in-house facility dedicated to the analysis of elemental impurities.

Measuring what’s there – and what it does

Our elemental impurities initiative is a three-tiered approach: i) quantifying trace metals in individual raw materials and in our final product; ii) determining whether our process of mixing and milling these raw materials itself contributes to impurities in the final product; and iii) understanding the impact of individual impurities on final protein quality. The first element was perhaps the most time-consuming, but the third part – understanding impact on protein quality – was the most important. To study this, we also utilize a CHO model system that produces a specific protein where we can assess the impact of elemental impurities on the ability and quality of the protein. The findings will not apply to all biopharma processes, but given that 70 percent of the biopharma industry uses CHO cells, the data is relevant to most systems.

One example of impurities I’d like to share is our learnings around ferrous sulphate. Cell culture formulations and cells in general, must have a source of iron. Ferrous sulphate is a common choice. Ferrous sulphate is sourced from mining from the earth and as such may have companion ingredients in the form of impurities. We found that our ferrous sulphate had very high levels of manganese, which in parallel the industry was learning had a high impact on protein quality. We changed our ferrous sulphate supplier to one that offered a product with a lower manganese level. To our surprise, however, one of our clients then reported a sudden change in the glycosylation profile of their product. Our subsequent investigations showed that this was caused by lower manganese levels in the medium, which was a consequence of our switch to a more pure ferrous sulphate. Essentially, the client’s process relied on manganese impurities for the required product profile. It was easy to fix with a manganese supplement, but the case serves as an interesting example of how a higher quality product can have an unexpected negative impact. (I might add that, for most customers, reducing the manganese impurity level was a positive development!) The whole topic really emphasizes the importance of understanding a product and its processes; it is unwise to rely on impurities for a bioprocess; far better to understand what the process requirements are, and then work with media suppliers to ensure those needs are met.

Another example of the importance of being able to accurately track down and quantify trace metal contamination also involved manganese. We were working with a customer to establish the source of a ten-fold excess of manganese in certain lots of the same cell culture medium. After many dead ends, we isolated the source of the manganese in the most surprising culture component: vitamins. Vitamins are synthesized using processes in which impurities are well-controlled, so this was unexpected. Nevertheless, a specific lot of vitamin B6 had up to 500 ppm of manganese, and was clearly the source of manganese contamination in the final product. Once again, we fixed the issue by working with our vitamin B6 supplier, and again it shows how a sophisticated trace metal analysis initiative can help identify even the most unusual problems.

Control and customization

At MilliporeSigma, our growing understanding of the effects of elemental impurities, and the expertise we have developed in the quantitative analysis of trace metals, has led us to focus on our raw materials and ensure that each individual material is as pure as possible. By ensuring that each component is high quality, we minimize the cumulative effect of impurities on the final culture medium and, hence, on protein quality. From the data we collect through our three-tiered approach, we continually modify our systems and guide ourselves to do things better, and to make better supply chain decisions.

Looking ahead, we have identified a market need for a customized trace metal analysis service. We are now in the process of developing this as a formal offer to help clients quantify trace metal impurities in their cell culture media. It also matches our philosophy of data visibility. Making data available to customers allows them to manage process variability according to their needs; for example, by mixing different batches of medium to ensure the cells receive optimal levels of given trace metals. Raw materials will never be 100 percent clean, but if you can measure impurity levels, you can manage impurity levels. Without a data-driven approach, it is far more difficult to control the impact of trace metal variability on bioprocesses.
My advice for industry is this: first, understand your particular process; second, communicate with your media supplier to ensure you build robustness into your supply chain; and third, use data to design a process that achieves the desired product profile.

Chandana Sharma, PhD, is Head of Cell Culture Raw Materials, Upstream R&D, at MilliporeSigma.