The viscosity problem

You don’t need to be a formulation scientist to know it’s harder to push honey through a syringe than it is water. It’s because honey is much more viscous, so it takes more force to drive it through and out of a needle. Liquid drugs that contain high concentrations of large and complex molecules, such as biologics, or that have molecules that can form bonds with each other, pose a similar problem: they’re often too thick and don’t flow easily enough to be delivered through standard injection devices. Typically, volume is added to thin out the liquid, and the drug is given intravenously in a clinical setting. This ensures the patient still gets the correct dose, but it’s also inconvenient for them to travel to the hospital and wait there while the IV bag drains. There’s the alternative of using an injection with a bigger needle that has a wider diameter, but this can be more painful – which might make people resistant to taking the drug, especially if they have to administer it at home.

To make delivery as smooth as possible for patients, many in the field are investigating how these drugs could be delivered at home via selfinjection. But for this to work, they need to figure out how to get smaller volumes of thick fluids through those devices while maintaining the drug’s stability. One approach is to alter the formulation in ways that reduce its viscosity, while another is to use a drug delivery device that provides enough power to push the drug through the skin. However, neither strategy is straightforward.

Reducing viscosity

The simplest way to reduce the viscosity of a fluid is to dilute it. But since the amount of liquid you can use for subcutaneous injections is limited – for comfort, the injection time should be reasonably short, while injecting too much can hurt – there isn’t much wiggle room where it comes to adding volume.

What you can do is add ingredients, known as excipients, to the formulation alongside the active pharmaceutical ingredient (API). While they’re considered inert substances, excipients interact with APIs in a variety of ways to help establish the properties a drug needs to be safe and effective, like improving its stability or extending shelf life. For example, an excipient could work to create space between protein molecules, which typically serve as the API within biologic drugs. This helps the fluid flow more freely, explains Shahid Uddin, senior director of formulation and stability at biotech company Immunocore.

Most proteins have a charge, which means they can be attracted to each other. This allows them to come together and form larger structures. “They become bigger, and with everything bigger it’s harder for them to move,” explains Uddin. “The excipient tries to remove that charge.”

But choosing the right excipient for the job can be complicated. Because every protein is unique and may behave differently at high concentrations and formulation conditions, there’s no one-size-fits all solution to improve viscosity, says Alana Gouveia, protein formulation scientist at Merck. The amino acid arginine is considered the standard choice for protein formulations, for example, but it doesn’t always work. Scientists must weigh up the properties of excipients alongside those of the API, and consider how each substance might behave within a particular formulation. There are plenty of factors at play here: for one, because the forces responsible for protein-protein interactions within highly concentrated liquids that can create viscosity are of the same chemical nature as those that keep an API intact, meddling with them can have negative knock-on effects. This is especially true when the API is an antibody, due to the susceptibility of these molecules to degrade through aggregation. “If you interfere with these interactions too strongly, we could also risk destabilising the antibody itself,” explains Gouveia.

Excipients can work synergistically when used in combination. At Merck, Gouveia and her colleagues discovered that using an anionic excipient with an amino acid was more effective at reducing viscosity, while maintaining protein stability, than using either substance alone.

However, excipients can only reduce viscosity to a point within a fixed volume, says Uddin. “There are some biophysical limitations, for instance, molecular crowding,” adds Gouveia. In formulations with a very high concentration, there isn’t enough space to prevent proteins from interacting with each other, so the impact an excipient could have is limited. “Another way would be to optimise the amino acid sequence of the protein while maintaining its biological activity and stability during the protein design, in the early development phase,” says Gouveia.

She adds that because the amino acid sequence of a protein may affect how it interacts with other proteins, including in ways that cause viscosity, you could alter the sequence to change that protein’s behaviour. This can also impact the protein’s efficacy, and as such, it requires a deep understanding of protein structure and properties to apply the technique effectively.

Delivery devices

For a delivery device to inject a viscous drug, it needs to generate enough force to push the fluid through a suitably small needle and under the skin within an appropriate amount of time and without causing damage to itself, says Michael Roe, director of device development at pharmaceutical company Kaleo. Achieving that force is the biggest challenge in administering these sorts of drugs, he explains. And to make things even trickier: when you apply force to the fluid, the pressure often causes its viscosity to increase – making it more difficult to push out.

Gas-powered devices are a possible solution here. Rather than using a spring to apply force, as seen with standard auto-injectors, the pressure is contained within a small gas cylinder. “The cylinder contains all the force, so you’re not putting the plastic under stress until the moment of evacuation,” Roe explains. When you puncture the cylinder, the gas expands into the interior of the device and creates a significant force on the plunger to drive the drug through the needle, he adds. And because pressure is applied more consistently than with a spring, delivery would be more comfortable. With simple spring-powered devices, more force is released initially when the spring is tightly coiled, creating a surge in delivery even at small volumes. “It still has an inrush at the beginning, and you notice that…it’s not a pleasant sensation,” Roe says of an injectable device he uses that delivers just 1ml.

Gas-powered methods may be able to administer drugs with a higher viscosity, but they do have some drawbacks. “As your operating pressures go up, the device has to contain all of those pressures, and plastic can swell, and it can act like a balloon. You don’t want it to explode,” says Roe. There are also limits to how much pressure can be applied for standard needle sizes. Another approach is to use on-body delivery systems. These are usually attached to the skin under clothing and inject the drug slowly over around 15 minutes to an hour. “The volume we can give is much larger,” says Uddin. Devices like this can be used at home, but they aren’t that common, he adds.

Other technologies use different means to create pressure, for example through a battery-powered pump or by using the gripping motion of the hand to generate force. Ultimately though, they’re all trying to do the same thing: generate enough pressure to get the drug out without breaking. “People have patented everything under the sun and they’ve all got disadvantages,” says Roe. “There are a lot of ways to skin that cat.”

The future of viscous injectables

As for the future of drug delivery, viscosity will continue to play an important role in the development of new devices, and Uddin expects to see the mechanisms that push the drug along the needle and into the target site to continue to improve. For players within this space, what patients want – ease of use, with as little pain as possible – will be of central importance, he adds. “Patients are now at the fore of everything that a lot of companies do, because that’s basically the customer, right?”

For Roe, what’s most exciting here is the work being done to enable longer injection times, which means that larger volumes can be given. This hugely expands the range of therapies that can be administered at home via hand-held devices. “It can carry a lot more convenience to the user,” he says. “A lot of these things would otherwise be done in an infusion setting or at a hospital.”

Over in the lab, digital tools such as machine learning applications could help speed up research by instantly evaluating the suitability of excipients or other solutions, says Gouveia. “It would reduce the time in the lab, and also material consumption.”

But for Uddin, the ultimate goal would be to not need injections at all. “The holy grail of biologics is for you to be able to put the drug in a cup and drink it, like you would cough syrup…there’s a lot of work being done in that area,” he says. “Can you imagine that you’d never need to have an injection?”