Why studies fail in week three

Infusion studies rarely fail at surgery. They fail two to four weeks in, when patency goes, a needle backs out, or an exit site turns. By then the animals are committed, the dosing window is half gone, and the choice is a compromised dataset or a rerun. A formulation that attacks the fluid path can end one sooner than that.

The failure modes

Seven things end preclinical infusion studies. All seven are decided at the protocol stage, by the hardware you specify before a single animal is enrolled. Week three is only when you find out.

Drug and vehicle incompatibility

What you see. The line stops working, or the delivered dose does not match the analytics, and the hardware looks fine on inspection.

Why it happens. Co-solvent and surfactant vehicles, strong acids and bases, and lipid emulsions all interact with polymers: softening or crazing of the catheter, swelling at a bonded joint, precipitate where drug meets lock solution, compound adsorbing onto the lumen wall so the animal receives less than the calculated dose, extractables entering the dose itself. None of this shows up in a bench flush with saline.

What to do about it. Run compatibility before the study starts, not after the first cohort is implanted. This is the sponsor’s to establish: a static soak and a flow trial on the actual assembly, with your formulation, at study temperature, for the study duration. We do not provide compatibility data for specific compounds, and no general material chart substitutes for testing the formulation you are dosing. Build the time into the protocol schedule.

Occlusion

What you see. Patency is lost and the line will not draw or infuse.

Why it happens. Thrombus at the catheter tip, fibrin sheath along the shaft, or drug precipitate in dead space. Risk climbs with every needle access and with every hour the line sits locked.

What to specify. Lower internal volume means less residual drug and lock solution sitting in the line. A design that lets you restore flow from either direction buys back a study that would otherwise be over. Ask what the dead space is, in microliters, before you commit.

Needle dislodgement

What you see. The non-coring needle backs out of the septum mid-infusion.

Why it happens. A rigid needle in a septum is held by friction and tape. An animal that moves, a jacket that shifts, or a tether that snags will eventually win.

What to specify. Remove the needle-in-septum interface and the failure mode goes with it. A flexible set that passes through the port into the vessel catheter cannot back out the way a needle can.

Dead space

What you see. Dose is wrong and nobody can say by how much.

Why it happens. Every microliter between the septum and the vessel holds drug from the previous access or lock solution from the last flush. In low-volume dosing this is not a rounding error.

What to specify. Put the port and catheter volume in your protocol as a number, the way you specify the dose. If a supplier cannot tell you the figure, that is the answer.

Exit-site failure

What you see. Infection, granulation or necrosis where hardware meets skin.

Why it happens. Any percutaneous component is a permanent wound. Buttons and externalized catheters trade surgical simplicity for a site that must be managed for the length of the study.

What to specify. For studies past a few weeks, a fully implanted port removes the skin interface. Where a percutaneous route is required, the anchoring and the material matter more than the price.

Catheter tip trauma

What you see. Vessel wall damage, thrombosis, loss of the vessel.

Why it happens. A squared or oversized tip in a small vessel abrades the intima. The vessel is lost, and in a long study you may not have a second one.

What to specify. Rounded, tapered tips sized to the vessel rather than to the catalog. Taper lets a larger lumen sit in a smaller vessel without the trauma of a uniform shaft.

Harness and tether loss

What you see. An animal defeats the system and the study arm is gone.

Why it happens. A single contaminated or chewed connection usually means replacing an entire harness, and the replacement is a second anesthesia the IACUC protocol may not allow.

What to specify. Modular hardware. Replace the failed component rather than the whole assembly, and keep the animal on study.

Talk to a surgeon, not just an engineer

Three of these seven are surgical, not mechanical. Tip placement, exit-site management and whether a model tolerates a second anesthesia are judgment calls that come from putting these systems in animals, not from designing them.

Our Chief Surgical Advisor, David Moddrelle, SRS, has thirty-five years implanting infusion and sampling systems across species. He will review your approach, your access site and your closure before you write the protocol, and train your surgical staff on the implantation.

Surgical services Who we are

The cost nobody puts in the budget

A device failure is not a line item on a purchase order. It is animals that cannot be used, a dosing window that cannot be recovered, surgical time already spent, and a reduction commitment to an IACUC that gets harder to honor. The hardware is a rounding error against any one of those.

Specify on performance rather than unit price. A port that costs more and keeps twelve animals on study to week twelve is not the expensive option.

Fifteen minutes before you write the protocol

We will look at your model, your species, your access site and your duration, and tell you what we would specify and what we would avoid. If the right answer is a device we do not make, we will say so.

No charge, no obligation, no sales call afterwards unless you ask for one. Twenty years building these systems, on a career that started well before AVABIO, means we have usually seen something close to your protocol.

Talk to an engineer Talk to a surgeon See the published record