The community of monoclonal antibody (mAb) producers and researchers continues to work hard to realize fully continuous production of mAbs. The main focus is on implementing continuous production using disposable equipment. Nearly all the necessary unit operations are available in continuous process mode today and, although there are still obstacles, such as the validated viral clearance step, major progress has been made, with successes at the miniplant scale (1). Most publications focus on fully continuous or semi-continuous processing using continuously operated equipment for the upstream part of the process, with the remaining processes being operated in batch mode. Economical comparisons of fed-batch and continuous processing have been published for the upstream part of the process only (2), (3). Hammerschmidt and colleagues focused on the comparison of a complete fed-batch process with a continuous process based on precipitation (4). But what about the cost of goods (CoG) between a typical fed-batch platform process and a fully continuous platform process for mAb production? Information in this important area is lacking, which is why we decided to perform a study to address the question: does the fully continuous production of mAb offer CoG benefits?
Our CoG analysis (5) considered the main process related costs, such as labor, capital, consumable, medium, waste treatment, maintenance, and buffer and media preparation costs (but not building costs). The initial base case scenario was set to an annual production of 200 kgAPI. The results? A fed-batch upstream process (USP) is more favorable than a continuous USP, but the continuous downstream process (DSP) is more favorable than a batch DSP. Specifically, within the upstream part of the fed-batch, as well as the continuous process, the fermentation medium costs dominate the CoG. Although the cell-specific perfusion rate was set to the lowest level published so far – 0.05 nL cell-1 d-1 (6) – the perfusion medium costs of the continuous process were much higher than the fed-batch fermentation medium costs. Overall, the analysis showed a CoG difference between continuous and fed-batch USP of 33 €/gmAb. Further analysis revealed that the continuous USP CoG stayed higher than the fed-batch USP CoG over a large range of cell specific productivities (20-90 pg cell-1 d-1) and perfusion medium prices (10–30 €/L). The picture only changed in the unlikely event of perfusion medium prices as low as 5 €/L. Therefore, fed-batch mode is more cost-effective for the upstream part of the process. Regarding the DSP, the continuous process mode was more cost effective than the batch mode by 8 €/gmAb. Within the continuous DSP, resins and filters were used much more effectively than batch DSP, leading to lower consumable costs and lower overall costs.
To conclude, the fed-batch USP and the continuous DSP were the preferred variants, which led us to investigate a hybrid process. The hybrid process consists of a fed-batch USP and a continuous DSP, which were connected through a harvest vessel. The hybrid process combined the advantages of both process modes; the hybrid process led to total CoG of 50 €/gmAb, whereas the fed-batch process CoG was 59 €/gmAb and continuous process CoG was 84 €/gmAb. The hybrid process stayed the preferred process mode within a wide range of capacities between 100 and 1000 kg/a. The fully continuous process could only be more cost effective if the cell-specific perfusion rate of the culture could be decreased below 0.017 nL cell-1 d-1 – a goal that can only be reached through the development of new types of media. In the future, fully continuous processes may provide a good alternative regarding CoG. For now, we believe the hybrid process shows great potential, and should be considered as a process mode for mAb production based on disposables. Our current research focuses on the successful demonstration of the hybrid process at miniplant scale.
Newsletters
Receive the latest analytical science news, personalities, education, and career development – weekly to your inbox.
References
- S Klutz et al, J. Biotechnol., 213, 120–130 (2015). PMID: 26091773 AC Lim et al, Biotechnol. Bioeng., 93, 687–697 (2006). J Pollock et al, Biotechnol. Bioeng., 110, 206–219 (2013). N Hammerschmidt et al, Biotechnol. J., 9, 766–775 (2014). PMID: 24706569 S Klutz et al, Chem. Eng. Sci., 141, 63–74 (2016). KB Konstantinov et al, Adv. biochem. Eng., 101, 75–98 (2006).
