In September 2006, Cleveland BioLabs Inc. approached SynCo with a challenge – could they help to develop a production process, scale up, and manufacture a cGMP batch of their recombinant protein treatment for exposure to radiation (Protectan CBLB502) in approximately a year? Normally, developing a new production process and scaling up from 5L to 1500L to produce a cGMP batch suitable for Phase I clinical trials, including filling, can easily take 18 to 24 months.
SynCo accepted the challenge and, as a result of truly diverse thinking, excellent communication and teamwork between CBLI’s consultants and its staff, plus sheer smart work and a ‘can do’ attitude, SynCo was able to deliver filled final drug product suitable for Phase I clinical trial use in October 2007, 13 months after the start of the project. There were definitely times when the deadline was in jeopardy, but innovative problem solving in close collaboration with CBLI and its consultants got the project back on track.
The timeline was challenging as a complete production process had to be developed, which in the end consisted of fermentation, primary recovery, two chromatography column steps, as well as an endotoxin removal step, followed by a scale-up from 5L to 1500L. Analytical methods were also transferred from CBLI to SynCo where they were qualified so that they could be used for batch release and stability studies. Following this, the cGMP production of bulk drug substance, filling of final product and its release was all completed in just 13 months.
So how did SynCo and CBLI rise to the challenge?
Before anyone put a pen to paper to map out the potential production process, CBLI’s Director of Chemistry, Manufacturing and Controls, Ed Venkat was keen that everyone involved was bought into the project and took responsibility for its eventual success.
“I wanted joint decision making and ownership all the way, as I saw this as the key to meeting our tight development timelines,” said Ed.
Ed joined CBLI after SynCo had been selected as CBLI’s development partner, but was more than comfortable with the decision because he knew of SynCo and its track record for high quality and meeting tight deadlines for process development and cGMP production.
SynCo and CBLI formed a steering committee, which agreed on weekly goals and follow-up action items despite only meeting face-to-face every six weeks. Its composition was a potential challenge in itself as Ed’s ‘United Nations’ team, as he called it, consisted of Americans, Russians, Dutch, Germans and British – all of whom thought and worked very differently.
However, Ed believes, “The cultural diversity of the members of the steering committee created a team dynamic that was a real benefit. The diversity of thinking made for excellent problem solving, especially when times got tough. There was ‘unity in diversity’ - both teams worked in unison towards achieving the common goal.”
Although the committee members agreed on all the key milestones, Ed was keen that, “These were not carved in stone, as everyone knew slippage might occur and it was important to be flexible and respond dynamically to any challenges on the horizon. We did, however, keep the milestones under review and set targets for each month to remain on track.”
The starting point
Right from the start, the omens looked challenging as Ed explains, “We handed over a process that had been developed in our research lab, which was extremely rudimentary and not well understood. The couple of grams we had made for animal testing were not well characterised.”
SynCo was confronted with an outline process that had been run in the lab, but had not been optimised and delivered very low yields. Clearly, this process would have to be modified and improved if it was to be scaled-up to yield a successful commercial product.
In addition to minimal characterisation, SynCo also encountered problems when they tried to switch from the research cell bank to the fully characterised cGMP Master Cell Bank needed for large scale cGMP production. They found that a small fraction of the cells no longer contained the key gene construct. This meant productivity would be low and there was a likelihood of a higher than usual level of host cell proteins, placing a heavy burden on downstream processing, which had potential to threaten the technical and financial viability of the project.
According to Ed, “SynCo used its considerable experience in E. coli fermentation to solve the problem by shortening the fermentation time after induction and, therefore, reducing the accumulation of non-producing cells.”
However, it took six extra weeks to develop a stable, manufacturing process that was scientifically sound and robust. This slippage placed even greater pressure on an already tight timeline for the full-scale cGMP production.
The challenge had just got a whole lot harder – and the downstream processing still had to be defined!
Getting down to it
As mentioned earlier, following fermentation and primary recovery, the downstream processing method used to deliver a suitable drug substance would involve two chromatographic steps and a final polishing step.
During development, CBLI had used an ion-exchange column as its first purification step, but the resins used were not suitable for use in cGMP production as these were not available at a quality compatible with cGMP production. In fact, many lab-based purification steps are later redeveloped to enable commercial production, as the raw materials used must be available at a quality compatible with cGMP production, i.e. from a stable supplier, able to provide it long-term. However, says Ed, “With SynCo’s experience in ion-exchange technology they had no problems in developing an appropriate step for Protectan CBLB502.”
Hydrophobic Interaction Chromatography - The real challenge begins
SynCo’s and CBLI’s development team’s mixture of experience and intuition led them to recommending hydrophobic interaction chromatography as the next step after ion exchange, despite it usually being avoided if at all possible, due to its variable, and somewhat unpredictable performance. However, it can deliver major benefits.
Explaining the thinking behind this apparently ‘high risk’ strategy, Ed said, ”Hydrophobic interaction was deemed necessary since it enables a true ‘step change’ in product purity to be achieved in a single stage without a great loss of product. In my experience, you can lose up to 20% of your desired product at each clean-up step. So, bearing in mind we would also need a final endotoxin removal step, this would make it a three-step purification process, giving a potential cumulative loss of up to 60% of our active product. Clearly, adding any extra steps was not a real option, so SynCo convinced us that we would need to harness the benefit of hydrophobic interaction, especially as it worked well in their preliminary evaluation.”
In the end, SynCo was able to successfully incorporate a hydrophobic interaction chromatography step, but defining and optimising the process put the project timelines further at risk.
In some cases, the use of two chromatography steps will reduce the endotoxin content down to an acceptable level, but in this case, a dedicated final removal or ‘polishing’ stage was included as an added insurance which had the potential to seriously reduce the yield. Fortunately, SynCo was able to use the latest technology and secure the purity needed with only minimal impact on the yield. According to Ed, “This final step was so well designed that it only reduced the yield by a mere 3%.”
Beating the clock
Despite the delays in defining the downstream processing, Ed was still pleased with the end result. “Yes, we fell behind on time, but if we had rushed things we would have soon gone past the breakpoint at which the commercial feasibility of the entire project would have been jeopardised. Instead, we jointly took our time to find the solutions that gave the correct balance between yield and purity at each stage of the purification process. Because of the pressing timelines, we were prepared to accept a yield that may have been lower than usual but we couldn’t compromise on the purity. In the end, we were more than pleased with the yield and this backed up my view that you have to look at the downstream far more closely than the upstream aspects of production, as this is where the true value can be added.”
Many contract manufacturing organisations would, at this stage in the project, run an intermediate scale production batch in their fermentors, prior to further scale-up to 1500L. However, due to SynCo’s extensive experience in large-scale production using E. coli, it recommended going straight from 5L scale to a 1500L run. This not only saved significant time, but also allowed enough material to be produced to allow the downstream purification process to be run twice during engineering run, ironing out any ‘kinks’ prior to the cGMP production run.
With the clock ticking inexorably down towards the original deadline for the release of final drug product, SynCo was determined to find ways to catch up on lost time. Its development team advocated finalizing development and preparing for cGMP production at the same time, a pragmatic but risky strategy. Despite this, the Steering Committee approved its merits and it helped claw back some of the missing weeks.
Furthermore, SynCo’s dedicated staff worked over the weekend and through the night to perform the tests necessary to give the final production process the ‘thumbs up’, thereby shortening the time between process approval and commencement of cGMP manufacture.
At all stages, SynCo ensured that the testing of CBLI product was treated as an absolute priority in order to cut testing times to an absolute minimum.
Two weeks out of fifty-two ain’t bad
In the end, the final drug product was released just two weeks behind from the original 12-month deadline, saving almost a year compared to the industry standard! And, it should be remembered that this deadline had been set when the problems with the master cell bank and the challenge presented by the use of hydrophobic interaction chromatography could not have been foreseen. However, as Ed said earlier, “The diversity of thinking made for excellent problem solving, especially when the times got tough,” and this helped both CBLI and SynCo prove that, in certain circumstances – 24 into 12 does go, but only just!