Talk to one of our experts
12 December 2023
In the drug delivery sector, time really is money. Every day spent developing a drug delivery system increases development costs and reduces the window between bringing a profitable drug to market and its patent expiring. Energy models are a tool which allow for the most critical elements of a root cause investigation to be targeted, potentially saving weeks on identifying and solving issues. Energy models can also be used to predict the performance of a device, and how this is impacted by design changes.
What is an energy model?
An energy model is a mathematical model that investigates the energy transfer at all interactions in a system. An energy model is advantageous to a force model as it is less susceptible to inertia effects and can more elegantly handle gearing effects and conversions between forces and torques allowing interactions to be compared on a like for like basis. As with force models, an energy model is based on how the system is designed to operate – it does not consider unexpected interactions.
Often, the majority of energy dissipated in a device is caused by just a few interactions. Therefore, it is common to see an energy model simplified to consider only the most significant interactions.
DPI development and specific challenges
To successfully reach the lower respiratory tract, active drug particles must be less than 5 µm in size; 1-5 µm tends to be the target range. Unfortunately, within this target range particles tend to be both cohesive and adhesive, with a marked tendency towards agglomeration. Developing an effective DPI that delivers particles of the required size, in the absence of any active device delivery mechanism, is therefore a substantial formulation and engineering challenge.
While carrier-free formulations are an option, a common strategy is to attach the fine active pharmaceutical ingredient (API) particles to larger carrier particles tens of microns in size. Lactose is a popular choice of carrier because it is well-tolerated by the lungs[3]. The API particles detach from the carrier during inhalation, which goes on to deposit predominantly in the oropharynx due to particle size[4]. With this approach, the carrier accounts for the bulk of the resulting formulation, making it easier to handle and to accurately dose the very small amounts of API required.
Formulators deploy a range of techniques to ensure dose dispersion and/or API detachment manipulating both the composition of the formulation and particle properties such as shape, surface morphology, and charge to achieve success. The particle engineering techniques deployed range from jet milling through spray drying and spray freeze drying to supercritical fluid technology, each associated with distinctly different physicochemical property development[5].
Testing early formulations in a prototype device is essential since product developers have the freedom to adapt both device and formulation characteristics to meet drug delivery goals. DPI delivery devices use complex internal geometries to translate the pressure drop induced by the inhaling patient into the energy needed for dose aerosolization, pressure drop across the device being a key differentiator for device design. Iterative development ultimately leads to a formulation device optimized for drug delivery to the lungs for a given API. Yet this is a complex, fine-tuning exercise[6].
Figure 1: Illustration of examples energies dissipated by each interaction, as outputs predicted by an energy model. Often, a few interactions dissipate most of the energy in a device.
How does an energy model save time during a root cause investigation?
There are several methods used to conduct a root cause investigation. One common method is to use a cause-and-effect diagram (also called Ishikawa diagrams). These diagrams help break down potential causes to a more granular level, which can then be tested to verify the root cause of the problem. For even a simple device, a cause-and-effect diagram can become very large. The cause-and-effect diagram shown below was produced for an inhalation device Springboard recently worked on. Over 50 testable potential root causes were identified, which could have taken several months to test and verify the true root cause.
Figure 2: A cause-and-effect diagram developed on a recent project.
To avoid an overly time consuming and expensive test program, an energy model developed for the device can instead be used to identify the most likely root causes and focus the investigation. In our inhalation example, a common potential root cause was for friction coefficients to be high, as there were several rubbing interactions, and inefficient lead screws in the device. Furthermore, inputs to the model, like coefficients of friction, could be varied to understand if variation in these parameters could explain the observed issues.
The graph below shows the energy dissipated by each interaction in the device, and how this changed when a particular friction coefficient was increased by 20%. The graph shows that the energy dissipated by the most wasteful interaction (‘CD’) doubled. This highlighted how sensitive the system was to changes in friction coefficient and demonstrated that this coefficient alone could plausibly explain the high energy loss they were observing.
Figure 3: Predicted energy dissipated for each interaction in the inhalation device (blue). The red bars show the increase in energy dissipated when the individual friction coefficients were increased by 20%.
Important considerations when developing an energy model
There are several important considerations to get right when developing an energy model, which include:
- Ensure that the team fully understands the device – Springboard has experience working with a wide range of devices in drug delivery and other markets so has plenty of experience gaining familiarity with new devices quickly.
- Perform tests to validate your model – Testing individual interactions is necessary to ensure that the model is capturing the correct physics. Springboard has experience with rapidly developing test fixtures for testing which limits delays with this step.
- Consider how the reality could differ from what the model predicts – could assembly errors or uncontrolled testing conditions be impacting the device?
What was the outcome for our client?
Prior to engaging Springboard’s services, our client had been testing one hypothesis at a time for over 6 months. Over the course of our 4 week project, Springboard was able to narrow down the large search space to only 3 potential root causes which required further investigation – saving our client time and money.
If you want to speed up your root cause investigation, please contact us today.
Written by James Alexander
Categories
- Drug Delivery (48)
- Med Tech (19)
- General (18)
- Human Factors (5)