Aseptic connectors do not assure sterile connections

Published in August 2024, by Adam Hiles and the Trisk team.

At Trisk, we produce large quantities of biologics by parallelizing and automating a fixed, small production scale. Parallelizing at scale reliably requires making many (automated) connections to and between components of our systems. It’s important that we do so without any meaningful possibility of contamination.

In order to to parallelize with minimal contamination risk, we decided to design and build our own connector and method for making sterile, re-usable fluidic connections.

The first question we’re usually asked about our connectors is: “why build your own?”

There are a range of reasons why the existing connectors on the market don’t meet our requirements. One is quite simple and broadly relevant to anyone hoping to make reliable sterile connections outside of a well-controlled aseptic environment: most connectors designed to maintain sterility throughout the connection process permit a route for microbial ingress. The chances of ingress are meant to be low, but they may not be as low as one might hope.

The need for connectors

Before we look at the connectors themselves, let’s recall the dilemma they’re meant to resolve. For successful upstream processing, we need a system free of contamination. Removing only life that constitutes “contamination” is not so easy, so generally we are less discriminating and start with a system free of all life, and with closed assemblies this is easy to achieve: many well-established techniques like steam sterilisation or gamma irradiation exist to create a sterile closed assembly. However, a sterile assembly is, by definition, unable to produce a cell culture. Nothing can grow so long as the system remains closed.

The need to reintroduce living organisms into the closed system necessitates some sort of connection to the outside. Within an environment designed for aseptic processing, such as an isolator, this connection can be as simple as opening the closed system and pouring or pipetting the desired reagents in. But when the bioprocessing system cannot be readily operated or moved into a suitable environment, such as for a large bioreactor in a clean production suite, more thought is required.

The key enabling bit of technology is the ability to connect two contamination-free systems together in a less-than-ideal environment without introducing any new contamination. When this is possible, we can:

1. Sterilise two closed systems, one of which is mobile and amenable to aseptic manipulation.
2. Open the mobile system in an environment suitable for aseptic processing to introduce the desired culture, but not any contamination.
3. Transfer the culture from the mobile system to the fixed (or less-mobile) system using the aseptic connection.

The standard aseptic connector design has a flaw

Many connectors purport to be this critical aseptic connection tech, and they largely follow the same concept. Here’s how they work.

1. There are two connector halves, each with a flow path on one side and an elastomeric seal, such as an O-ring, on the mating face.
2. Each elastomeric seal is covered with a polymer membrane that is affixed (normally heat sealed or ultrasonically welded) to the seal face and doubled over, allowing the seal to be removed without rubbing “dirty” membrane over the sealing face. The membrane doesn’t slide, it peels. But not yet!

3. An assembly, including the connector with elastomeric sealing face and membrane covering, is sterilised as a complete unit. The connector new forms part of the boundary of a closed, sterile system.
4. To make the connection to second closed, sterile system, the two connector halves are pressed such that the elastomeric sealing faces pinch the 4 layers of membrane between them. Typically, some mechanical locking occurs to hold the two connectors and two membranes together
5. The membranes can then be peeled off the seal face with the elastomeric element sealing around the membrane as it is removed, uniting the two closed systems through the connector and (hopefully) maintaining sterility, as shown in an exaggerated fashion below.

It is clear to see from the (exaggerated) images a fundamental issue with this connector style: the continuity of the closed sterile boundary during connection is fully reliant on an elastomer deforming around 4 layers of membrane in a tight enough manner to prevent ingress of contaminants.

This is problematic. Even basing the numbers on an impractically thin membrane sheet like clingfilm, the four layers would lead to a separation of ~60 µm — more than 2 orders of magnitude greater than the pore size of a sterilising filter. With a more viable membrane material you would expect a separation more than 3 orders of magnitude in excess of the sterilising filter pore size.

It is possible to check the reasoning made above in the real world with a simple experiment. A simple connector mock up can be made from bonded acrylic sheeting, standard O-rings and a membrane. Fastening two connectors together in the “mated not pulled” state we can then apply gentle (<0.2 bar) air pressure, submerge, and perform the connection process. A lack of bubbles is indicative of the integrity of the connection. However, if at any stage appreciable air can get through, it’s safe to say microbes can too.

In the tests shown below, the mating was paused pre, mid, and post pull to show representative points of the connection. The test jig had the minimal 60 µm separation (think clingfilm membrane). As you can see, even with the thinnest conceivable membrane, we will not be passing a bubble point test in the half-pulled state!

When aseptic is not good enough

Aseptic processes aim to limit the introduction of harmful contamination to an acceptable level, and in many applications the transient open system produced by the standard aseptic connector may not cause an issue. Many processes have an acceptable failure rate greater than that induced by the connector, and these connectors would not exist if they weren’t well adapted to many uses. However, if the aim is to get it right every time, these sorts of connectors may not be good enough. Challenge test data from Merck found that 6 of the 7 connectors they tested failed to provide a sterile flow path in a contaminated environment. That some of these connectors are referred to as “sterile” connectors, rather than “aseptic” connectors, is a bit of a misnomer.

Whilst “aseptic” refers to a method or process that prevents the ingress of microbes, “sterile” is a state in which all life has been actively killed or removed. Connectors that create true sterile connections do exist, including the one we have patented. These connectors form a third, effectively closed system that envelops the connection point and can be sterilised-in-place, actively killing or removing any microbes. The connection can then be consummated in the newly formed sterile — not just aseptic — environment. When sterility is required, there’s really no substitute for sterilisation.

Of course, the extra assurance that comes with sterilisation requires making tradeoffs and accepting new constraints, and when we look at what’s available off the shelf, it’s not hard to see why everyone in the industry accepts the aseptic status quo, regardless of its faults and limitations. But at Trisk, the ability to make truly sterile connections was a problem worth solving — important to making perfect AAV at every stage and scale — and so we’ve gone beyond the shelf and found a set of tradeoffs that are well worth making. Stay tuned for more on that in a future post.