Pharmaceutical “specials” manufacturers are called upon to produce a wide array of medications in small batches and at relatively short notice. Some of these specials such as insulin can be very difficult to produce and control the quality. As with most proteinaceous solutions, insulin adheres strongly to plastic meaning that concentrations and chemical ratios can changing during manufacture, depending on the contact with plastic. However, there are several solutions to get around this and guarantee the quality of the final product.
Why we shouldn’t consider removing plastic from the workflow If proteins adhere to plastic isn’t the best solution simply to remove plastic from the workflow? Well, technically, this would help but it is about as impractical as it gets. To start with formulation processes are highly standardised and hard to change. There is a very good reason for this. The formulation process requires extensive testing and repeating to make sure that it works for every single batch, changing this regularly would take too long and run the risk of errors creeping in. Secondly, plastic is an incredibly useful material in manufacture. In fact, in the case of insulin, it is often delivered in pre-filled plastic syringes because this is the standard that hospitals require. Removing plastic and replacing it with other materials would be costly and likely introduce new problems. The final reason is that in most cases, insulin and other proteinaceous medications are only a fraction of the other specials that are manufactured at any given site. Making significant changes to the manufacturing process for one batch would significantly increase the time and cost of production, as well as reducing the flexibility of the manufacturing to produce a variety of specials.
Why we should do more testing instead The key to increased confidence and quality when manufacturing specials like insulin is better testing. This is especially important in specials production where testing of every single batch is not strictly mandated by regulators. In these cases, many specials manufacturers are taking matters into their own hands by implement Proactive Quality Assurance. This means testing more than is required by regulators in order to have better control over quality standards as well as to preempt regulation changes in the future. The problem with testing using gold standard methods such as HPLC is that they are expensive and can take a long time to get results, meaning that batches of medication have to wait a long time to be released for the patients. However, testing each batch with difficult formulations like insulin is the only way to be completely confident that the concentration in the syringe is the concentration that we expect it to be.
In order to implement widescale testing of specials for every batch, we have to find a simpler method that matches perfectly with the gold standard. We recently published an article about how and why it is a good idea to introduce new testing methods in the quality control processes for specials. Here we will outline exactly how you would do this for a particularly difficult case like insulin.
Which method should you use to test insulin For the sake of speed, cost effectiveness and ease of use, we are looking for a complementary method to HPLC. We use absorption spectroscopy as the basis for our DrugLog® instrument which can give accurate readouts of the concentration and composition of liquid drug preparations. It is also significantly cheaper and easier to run than HPLC meaning that specials manufacturers can perform testing at different stages of production, for example when the formulation has just passed through a mixing step in a plastic container. From day one, the DrugLog® will allow manufactures to compare the consistency of batches, relative to one another. This is extremely useful for detecting significant deviations between batches and raise the alarm that something is not quite right. However, relative differences are not nearly as useful as being able to accurately determine the final concentration of a specials formulation. For this, standaradisation and benchmarking are required.
Benchmarking absorption spectroscopy against HPLC The fastest, cheapest, and simplest method in the world will be no good to specials manufactures unless it gives exactly the same results as HPLC. This is the gold standard and it is imperative that any new methods match up.
Fortunately, it is easy to benchmark absorption spectroscopy methods like the DrugLog® against HPLC by running a detailed internal calibration. To do this, both the DrugLog® and HPLC must be used in parallel for a period so that results and measurements can be compared. If any deviations occur, the results from the HPLC can be used to further calibrate the DrugLog® so that the accuracy will steadily improve with each passing test.
Prerequisites for good calibration There are many ways to calibrate instruments but based on our experience, there are a few important prerequisites to get absorption spectroscopy as close to the HPLC measurements as possible.
Firstly, never use anything less than a 3-point calibration. Take a low concentration, a high concentration and an intermediate value and measure all 3 using both HPLC and absorption spectroscopy. This will not only improve the accuracy but also save time as it will help to quickly define the relationship between different concentrations and their respective readout for each instrument. The more points used in the calibration the better so a 5-point calibration would be preferable but, in our experience, not strictly necessary.
Secondly, it is essential to always calibrate using samples from real batches. This can sometimes be a source of frustrations as calibration inevitably takes longer but the payoff is worth it. The problem with reference samples prepared in the lab is that they are never exactly the same as a production sample. If we come back to the way that insulin interacts with plastic, it is easy to see why. It’s impossible to accurately replicate in the lab all of the plastic interactions that an insulin solution would have during production, meaning the calibration results would be less accurate.
The end result After repeated calibration, there will be a natural point where the HPLC results and the absorption spectroscopy instrument will always match. At this point, no more calibration is necessary, and it is possible to get accurate results using a cheaper and faster absorption spectroscopy method like the DrugLog®. This means that you can now fully control every element of the production process by repeatedly measuring at different stages. Now your production facility can be confident in their processes and their product and they will have a large amount of data to show both customers and regulators. Does this mean the end for HPLC? Absolutely not. HPLC is still the gold standard and will be for the foreseeable future. The goal is not to stop using HPLC but to implement additional testing where HPLC may be too time consuming or costly, ultimately giving specials manufacturers more control over every batch. This means that HPLC should still be used regularly to confirm measurements for absorption spectroscopy, but this will be more cost effective in the long run as it will be possible to drastically reduce the amount of HPLC test that are carried out. The end result is that as time passes, you gain more and more confidence in each method and the quality of your batches as you now have a failsafe in your quality assurance process.