In his latest column, Sarfaraz K. Niazi, PhD, went over some best practices for implementing innovative solutions for the purposes of streamlining biosimilar development.
The phrase "Don't fix it if it isn't broken" is an idiomatic expression suggesting not to make changes or improvements to something functioning well or with no apparent problems. The phrase became popularized in the 20th century, particularly in engineering and maintenance. It highlights the idea that spending resources or effort on something already functioning properly is unnecessary and potentially wasteful.
The exact origin of this phrase is unclear, but similar sentiments have been expressed throughout history. One possible origin can be traced back to a proverb that dates to the 17th century: "Let well enough alone." This proverb implies that it is better to leave something as it is if it is already satisfactory, emphasizing the value of stability and the avoidance of unnecessary changes when they are deemed unnecessary. I was taught this lesson by Thomas Hodgson, the president of Abbott International when I quit my tenured professorship to join Abbott International. He was adamant about being practical—not perfect—as he saw many imperfections in me.
The phrase applies well to biosimilars that must be “highly similar” and not better. This is a difficult path to tread, a narrow lane to traverse, with little freedom to drift, requiring staying on the course. Biologics are complex and consequently tricky to emulate, as large as over 1000-fold the mass compared to aspirin. Though they all come from a known and confirmed primary amino acid sequence, how this sequence folds into a 3-dimensional structure and creates domains that bind to receptors, remains a paradox first mentioned by Levinthal; even the arrival of artificial intelligence (AI)-driven protein structure predicting algorithms like the AlphaFold2 and ESMfold are still struggling to be accurate.
It has been calculated that up to 100 million variants could potentially exist for a monoclonal antibody. All this makes for complex challenges when comparing the structure of a biosimilar candidate to its reference product. One batch of the reference product will differ from the next, and sometimes the difference can be significant following a change in the manufacturing process.
Novelty is significant in fostering creativity, driving innovation, attracting attention, and pushing the boundaries of what is known or expected. It brings freshness and excitement to various human endeavors, leading to progress, discovery, and a richer understanding of the world. In the scientific world, novelty is associated, almost invariably, with something that is most desirable.
However, when it comes to biosimilars, reverse teaching is needed. Any novel change in the manufacturing process is most likely to change the molecular structure, pre- or post-translational. Whether this difference from the reference product is relevant is a moot point—it will inevitably lead to extensive testing, most likely leading to sizeable clinical efficacy demonstration in patients if the differences are minor.
The chance of significant differences can quickly kill the product. Such novel steps are welcomed when a new product is developed, not when a biosimilar is developed because there is no motivation. The cost of goods of biosimilars is very low since it only depends on the carbon cost, unlike chemical drugs, which may cost a lot more to produce. Every antibody can be made at much less than $100 per gram. At this cost, the packaging cost may even be higher than the cost of the drug substance. Why then bring changes to the process?
Recently, I wrote about the FDA’s final guidance on the continuous manufacturing of biological drugs, drawing my readers' attention. However, adopting this process when the reference product is manufactured using a batch process is not advised. The same holds of the most misconceived suggestion to bring in cell lines with higher protein expression without realizing that pushing a living entity can only result in an aberrant outcome.
The yield of a recombinant cell line is manipulated by various optimization strategies, including genetic modifications, cell line engineering, optimization of culture conditions (such as media composition, temperature, pH, and oxygen levels), and process improvements, all of which can impact the quality attributes.
Therefore, it is better to have a low-expressing cell line that will provide consistent and matchable characteristics. The only saving that comes with using a high-yielding cell line is the bioreactor size. The media cost remains almost identical since the carbon feeding produces the protein. There are many novel ways of using a smaller-size bioreactor and avoiding scaling up by combining the yields from multiple vessels, likely the single-use ones. My patents, US-9587283-B2, and other may allow this arrangement, and this patent is in the public domain. This is a much better way to reduce the cost than the manufacturing process.
Process changes are inevitable and desirable but do not have novelty steps. A good example is etanercept, which underwent 3 major revisions to the original: replacement of non-irradiated with γ-irradiated serum, adding an extra purification step to decrease process impurities (protein A ligand), and implementation of serum-free cell culture. The cell line manufacturing etanercept reference product yields less than one gram/liter.
Another challenge that comes to biosimilar developers is to choose a different cell line than used by the reference product. The FDA does not object to using another cell line, which brings many challenges in matching the post-translational modifications. This challenge has become more significant since initially, mouse myelomas (NS0 and SP2/0), baby hamster kidney (BHK-21) cells, and human embryonic kidney (HEK-293) cells were acceptable as hosts for recombinant protein production, but now the focus has shifted towards the Chinese hamster ovary cells. This is not driven by novelty; it relates to the product's safety. However, choosing between bacterial to mammalian systems is not advisable, albeit allowed.
The novelty in developing biosimilars should be focused on assay technology and a greater understanding of the impact of variants on efficacy and safety. Also, AI-driven approaches to data analysis, experimental designs, and statistical modeling should also be promoted to justify smaller clinical studies.
Novelty claims can draw attention, and they always work when they can bring a paradigm shift, but when the challenge is to walk a narrow trail, such as in the developing biosimilars, a novelty can bring much disappointment, and should not be introduced for the sake of inquisition, following what Thomas Hodgson said, be practical, not always perfect.
Hassett B, Singh E, Mahgoub E, O'Brien J, Vicik SM, Fitzpatrick B. Manufacturing history of etanercept (Enbrel®): Consistency of product quality through major process revisions. MAbs. 2018;10(1):159-165. doi:10.1080/19420862.2017.1388483. Epub 2017 Nov 29. PMID: 29020515; PMCID: PMC5800370.