Successful Methods for Perfusion Process Optimization
Welcome to the cell culture dish podcast successful methods for perfusion process optimization. I'm brand new surgeon editor of the cell culture dish joining me. Today is dr andreas custodian. Dr kostov on is principal scientist for g._e. Healthcare's life sciences business in this role andrea provides leadership and support to bioprocess business his main area of expertise. This is in upstream processing prior to joining g healthcare. Undress was director of upstream development at swedish orphan bio veteram a._b. Working with development of expression systems process development of microbial and mammalian cell based processes and scallop to see g._n._p. Manufacturing scales during the last twenty years andrea has held several positions within biopharmaceutical development including project and line management as well as manufacturing. I wanted to start today by asking what makes perfusion good manufacturing platform for me. There are several reasons to to go for fusion perfusion allow short residence time in the reactor continuous addition of fresh medium and the removal of service and that's makes it of course suitable for unstable q. <hes> perfusion allows to run close assise also with with high magic productivity by maintaining heist okay so densities over a long time with many opportunities for process intensive case and perfusion can be a convenient way to quickly get get started with your production of material given that you have a good batch medium. The sounds like are there specific product types or situations that are in especially especially good fit for perfusion. Oh yes yes. They're all <hes>. I see three -arious. One of course is the production of of unstable products this as mentioned inching before <hes> then there is integrated continuous manufacturing platforms. That's leverage the high of the metric productivity and the continues harvest upstream is connected to continuous downstream operation. Finally there's all types of process intensification intensification for removing bottlenecks <hes> that could be in southbank manufacturing to produce high for them to sell things it could be in the trains range to a bit <hes> c. train stages all produce the inaugural on and finally it could be hybrid prophecies between and fusion and with respect to a profusion processes <hes> could you describe the importance of optimization and then what are the various approaches for optimization. Yes optimisation is of course very important to gain the product quality and the process they call them. Be that that you're aiming at the soccer. Medium here is of course the most important component or device in order to get good throughput in the optimisation it needs is to be performed in a in a state of however. They'd results needs to be verified by a reactor conditions that could be then in the in an amber or as we use the the waves twenty-five system. I saw a recent poster that you presented at in in this poster you you discussed the development of perfusion specific media. I'm wondering if you could discuss a little bit about designing medium that supports low oh a cell specific perfusion rates and also could you explain what cell specific perfusion rate is and why a low one would be important. Yes i want to start with explaining what specific perfusion rate is. It is the volume of of medium that is at its it's pastel and <hes> so let's assume you you have a medium that supports a specific perfusion rate and i i call this the c._f._p. Are all of fifty and if you want to run the process at thirty billion cells per hour you'll metric perfusion rate that the bio reactor autumn's today will be one point five. If you want to use the the same medium <hes> and from the process at hundred million south timeout you will need need to have a volume metric few rates of five reactive audience today and that is of course <hes> unfeasible in production situation assuming that you have <hes> a medium with a better depth and which allow us to to to run of south pacific fusion rates of ten not talking silent they the same cell densities of thirty million south prairie my other result and we'll point three react up all day hundreds million sounds that's promo would result in one wrecked of today and that is of course <hes> a feasible sustainable manufacturing process and i'll i'll recent work we investigate it's screeing methods to develop media that allow a very low e._s._p._n. I'll starting point the the the based on medium and we screened <hes> different feet solutions that we call cell tooth and bech smolts the south boost with a positive impact with i'm taking to me study and twin spin tooks that is up pseudo perfusion conditions the result it and from relations that were then taken to iraq for fusions in wave or in the the d._r. That's really interesting and thank you for explaining that. I had another question. This is something that comes up. I think a lot in terms of media in general and that is what if you already have a medium that your cells are performing well in. Can you turn a fed batch medium into a profusion decision medium. Yes transferring a fat batch into the fusion process. Something i got many many questions around the they methodology that that that i described on the poster works actually very nicely for taking fetch meet with with its seats <hes> into perfusion medium and by the way we we also published a paper on this topic a month ago when biotechnology progress together with a with a coop commune <hes> where we did this for two cell lines and three setbacks media and and and so it's it's it's definitely a something that that can be easily be be accomplished with the with the methodology described and i just wanted to say for our listeners. I will have links to to the poster as well as i can get a link to the paper and post that in the show notes as well so you can easily find those references that addresses referring to how did the methodology work in the case study that you presented in the poster. The first thing i want to mention is it's a very fast method. The bench screening takes basically owned the only one week and the suda perfusion <hes> design experiments study two weeks so so within a month. You have the composition for a for a fusion medium that fits your your clone own. That is your production line. <hes> we use these smith all the chief for for two media active pro and and c._d._m. Finance zero so <hes> for our internal has been producing that line both media where where in bi-reactor fusions and foot boasts us media <hes> we could cheese <hes> c._s. biaz below twenty peak elites up a day and when we were pushing one the we could put even get down to just haven't peak liters per day at the same time <hes> the south pacific productivity was comparable talk to us that we were seeing in chicago so <hes> pushing down the the amount of media profuse the iraq that taught compromise the productivity when of the other things i really liked the poster was the case study used small scale models and then entrance for the process to a fifty liter stirred tank bio reactor. I was wondering if you could tell listeners. How predictable was your chosen scaled down model. We use they develop media from from the spin soups to perfusion by reactors well the the waves twenty five has been all working horse so the waste twenty five <hes> allowed us to to reach more than two hundred million themselves per firm out in addition to the p._d. Scale for fusions. We also stayed up to the to the next fifty <hes> bi-reactor with with t._f._f. T._f._f. Spin to model was suppose very good for predicting most partly. That's like the peak these <hes> <hes> reached the titus q p growth rate c._s._b. Are the glucose lactate metabolism off malaria and product quality like like charts variance and accurate gates <hes> however the model was not perfect doc <hes> for predicting the thomas the tank mates <hes> met mataafa missing even for the like in distribution <hes> <hes> we saw some subtle differences between the spin shoop and the p._d. Escaped by the way for the the let let later the p._x. P._r. and then how predictable was your performance between your process development and fifty liter scale the performance foreman between the the twenty five of the fifty perfusion boss was very good for for actually <hes> both <hes> gross relate late to <hes> metabolism related and finally protrude. I wanted to kind of switch gears a little bit. We're talking a lot about media. The optimization but another approach to optimizing perfusion processes is to optimize the equipment used in the process in another poster that you present a sack you present the use of hydro cyclones for cell separation and i was hoping you could describe a little bit about how hydro cyclones work why they make an attractive active alternative to traditional separation devices. Yes yes definitely <hes> hydra sykes on is a device comprising the thing all the cylindrical and conical part enabling essentially fuego separation provided by the by the feet suspension shen introduced tangentially at high flow rates into the into the device so the the absence of of rotors or moving parts apart from its compactness makes us an interesting alternative to confucian long term operation so now where it's a simple device <hes> that smoked prone to clogging or seething effects that is usually seen with with sedation based devices. What were your experiences with the hydra cyclone in your case study <hes> we investigate that separation efficiencies from fees at a different slow rate to some concentration and integrations of all of the hydro cyclone with the bio reactor. We found that's all seventy to eighty percents could be achieved with the current hydro sites on define. We run a fusion process in an extra fifty buyer react at fifty million south. I'm out faw for more than two weeks <hes> in in a very stable conditions sean so <hes> it was a very very successful <hes> test both of the device yeah thank you. I thought it was really really interesting to read the poster. We've talked about several different optimization methods but i wanted to come back and summarize a little bit if companies are limited time and the resources for optimizing their perfusion processes. Do you have some advice on what you would focus on. I it's it's definitely the medium that you should focus on <hes> as explained is this e._s._p. Are or or a low see s._p. Are the key to reaching high voter metric productivity low. Hello hello metric fusion rates and high product concentrations and as a result good process economy then you can consider consider what could be a relevant production scale and how you could leverage profusion to decrease the production skate and the biotechnology pro. Oh chris publication. I mentioned we showed that the waves twenty five college quickly operated for almost thirty days at v._c._d.'s between eighty eight hundred million south sounds program out in the largest southbank at working volume over twenty five liter. The reactor can easily live about forty grams of maps <hes> today. Thank you so much. This has been really interesting and i think profusion is such a important manufacturing process ticket sitter and moving forward. I expect we'll see more and more perfusion processes in the industry. Do you have anything else that you'd like to add. Sorry mr before we close for the day yes before moving into perfusion. You should ask yourself. What makes most sense for you. What does your infrastructure. What are you prerequisites. What is your platform and previous knowledge. Then you can decide where perfusion makes most sense in your process whether it's in south manufacturing c. train or your production by reactor the medium of stem the most important factor to work with. I hope i have given you the feeling that it is not difficult to use existing medial fell to develop a fusion medium that fits your your clone in this this message however is adding new components <hes> all the time and making the medium richer and richer before for moving into a preclinical production i would recommend to discuss the medium with your high loan application specialists to see the of components can as well be removed that just to decrease for example majority but also discounts the manufacturer ability of the medium for for the gala but yes thank you so much this as i mentioned it's been a really interesting podcast and i really appreciate. I think these are a lot of questions that people have about optimizing perfusion processes and i think it's a really good information to get out there that you can take f._a. I batch media and make it into a perfusion media and i think that that's really important so as i mentioned all of the posters we discussed discussed and the paper we'll have links to those in the show notes so at people can access those easily. Thank you so much. Thank you again for your time today. Thank you very much for having me. Thank you for joining us for this edition of the cell culture dish podcast to learn more about this and other stem cell bio manufacturing related topics. Please visit us at w._w._w. Dot cell culture dish dot com or for downstream bio manufacturing topics w._w._w. Dot downstream column dot com.