Companies commission us to:
Independent of suppliers and contract manufacturers (CMOs). Consulting and implementation.
Process development is the first step from laboratory scale to industrial production. Solid foundations in process development form the basis for the future of your product! Process development and implementation in production is usually carried out by a specialized department ("MSAT" Manufacturing Science and Technology). Alternatively, you can outsource the entire process to us.
We are independent of suppliers and contract manufacturers and can effectively represent your points of view.
This saves you valuable time and gets you to market faster!
In process development, trials and tests are still carried out on a small scale. In other words, trials in the laboratory or in the pilot plant, as an initial scale-up of production. Trials and changes to parameters can be carried out very quickly. Even series trials are possible: data can be obtained quickly.
And at a fraction of the cost of a later production trial directly in industrial production. This data is also used again and again at much later stages of the product life cycle. For example, for risk analyses, validations or subsequent process changes.
Many roads lead to Rome!
Depending on the company philosophy or market requirements, the requirements for the same product can be very different.
For example, if one company strives for maximum process throughput and minimum production costs, the priorities of another company may be completely different: batch sizes that are as flexible as possible, for example, or a particularly low impurity profile of the active ingredient.
At the end of a development phase in the laboratory, you have a lot of data. The problem is that this data is unstructured. You have to fight your way through numerous reports, tables and memos to extract the data.
If data is obtained with statistical test plans, we can easily convert it into mathematical models. These data models are the most compact form of representation of all the tests carried out. We calculate simulations based on these data models in order to be able to make predictions even before implementation in production. A significant gain in time!
Traditionally, process optimizations are carried out in the production process. Such tests are very unpopular as they involve significant financial outlay and a high loss of time.
Process optimizations on data models of key process steps are better and more useful. We use these "digital twins" to create realistic forecasts for the possible effects of process changes. This cannot completely replace trials in production. However, it does save many failures. This digital transformation of manufacturing processes is the most important lever for future profitability!
When transferring from laboratory scale to industrial reality, size changes by a factor of 1000 are often skipped. In the laboratory or pilot plant, for example, gram quantities are produced in order to carry out initial clinical tests.
In industrial production, however, we are in the kilogram range, with correspondingly modified systems. Suddenly, mixing processes are no longer trivial (e.g. with shear-sensitive cell suspensions) and temperature gradients are suddenly no longer negligible in larger reactors.
Here we draw on the data from statistical test planning and data modeling. We cannot avoid all "surprises", but we have the highest possible forecasting reliability.
All in all, this is a much better starting point than a risk analysis, no matter how comprehensive. Risk analyses only represent estimates of the risk and do not replace hard data
People often try to shorten laboratory tests to save time and money. Although this is true in the short term, there is no getting around the generation of data. Reliable data can only be generated with planned and executed tests.
For a certain number of parameters (such as temperature, time, pH value, etc.) there is a minimum number of tests that cannot be undercut - if you want statistically reliable data. You only have the choice between tests in the pilot plant (fast and cheaper) or in production (expensive and very slow).
Whenever possible, we generate data in small-scale tests and then transfer it to production using our data models.
We create a robust production process based on data. In order to keep the testing effort within reasonable limits, we carry out statistically planned tests, which we map in data models ("digital twins"). With these simulations, we achieve fast and reliable implementation in production.
In the first phase, clarify which conditions should be met for future production:
Flexibility, throughput time and batch size in particular form the basic framework of a future process design. These three parameters determine how you will later produce optimally:
The most efficient and effective way is to summarize data from the laboratory and pilot plant in "models". These models are mathematical descriptions of the production parameters and how they affect the quality of their product.
Data quality is essential here. In order to develop good models, you need data from so-called "balanced designs". These are special types of statistical experimental designs that vary parameters systematically and symmetrically. This is the only way to ensure that your experimental design does not already favor a certain result. A concrete example:
If you vary the temperature in many steps of a chemical synthesis, but hardly vary the pressure and pH value, you will probably come to the conclusion that your synthesis is mainly dependent on the temperature.
However, if you carry out a symmetrical variation, i.e. temperature, pressure and pH value are varied the same number of times, you will achieve a correct result. Only then will your model correctly reflect your "result space".
With correct models constructed from systematic tests in the laboratory and pilot plant, you can optimize your process settings even before they go into production. The usual questions that are clarified here are:
As you can see, the company philosophy naturally plays a decisive role here! A process design for the lowest impurities usually goes hand in hand with the most stable process conditions, but is often not identical with the highest yield.
This is where the company philosophy comes into play! Do I want the highest yield and the lowest process costs or the product with the highest quality and the fewest impurities?
Once the optimization and simulations have been calculated, the next step is to implement the industrial process. Even the best simulation only gives a possible result. The truth can only be determined through production runs. However, if the process development, optimization and simulation have been carried out properly, you will usually obtain goods that meet the specifications from the very first industrial batch.
You are not yet validated, but you can go straight on to validation with possibly only very minor adjustments. Marketable material is then already produced there.
Such a systematic approach has very strong financial advantages! Failed production runs on a full production scale cost many times more than laboratory and pilot plant trials.
There are enormous financial and time implications here! We therefore always carry out laboratory and pilot plant tests before implementation in industrial production. This is the only way we can achieve extremely short project durations and very short implementation times!
The challenge in process development arises mainly from the conflict between laboratory resources, project plan and industrial production. Very often there seems to be no possibility of carrying out laboratory tests or pilot plant tests according to a test plan. The argument here is often based on costs, time and lack of resources.
However, the correct process parameters are also very rarely guessed correctly. It is usually the case that the problems only really become apparent when they are implemented in production during the first technical production trials. Of course, the "hat is on fire" at the beginning and suddenly resources and costs are no longer a problem. The only problem that then becomes much more dicey is time.
We advise all our customers to invest time in laboratory tests rather than in the countless problem-solving meetings ("who's to blame?") that become necessary after faulty productions
Take the time to carry out well-planned trials. The results you achieve will make your life much easier throughout the transition to production.
We are happy to support you in test planning and execution as well as in implementation.
The goal of manufacturing process development for the drug substance is to establish a
commercial manufacturing process capable of consistently producing drug substance of the
intended quality
