Note concerning In-Vivo versus In-Vitro Friction

(of coaxial catheter systems)
(by J.J.G. Bos)

Here follows a short discussion about the discrepancy often encountered between clinical performance of coaxial endovascular catheter systems, and that of an in-vitro (engineering) model. The reader might benefit from extrapolation of the described effects to his or her own field of medical device performance modelling.

Please click HERE in order to learn more about friction properties of the EMPEH coating.

Endovascular catheters have the advantage of entering the body through a small hole in the skin, thus allowing diagnosis and intervention without a full-blown surgical operation. The catheter has to follow path that is dictated by the shape of the chosen blood vessel, which sometimes leads to quite a lot of friction. A floppy catheter will easily comply to the vascular curvature, but its lack of stiffness makes it hard to push it through. Modern endovascular catheters therefore show a well-balanced design that optimizes the total of all relevant functional parameters such as stiffness, torque, shape memory of the material, and outer diameter (size).

Coaxial catheter systems are quite hard to operate when it comes to friction. A catheter that is slid over a guide wire sometimes encounters so much friction that it eventually gets stuck before reaching its end point. This phenomenon is referred to as "in-vivo guide wire friction". The same happens sometimes in the case of one (smaller) catheter being pushed through a larger guiding catheter (GC). ("in-vivo catheter/GC friction").

When tested in vitro - for instance with water or saline at body temperature, and in a curvature that closely resembles the vascular situation - experimenters often find that friction is of no issue whatsoever. The catheters negotiate the curvatures without encountering significant friction. This is often in stark contrast with above-sketched the in-vivo results.

A qualitative impression of the in-vivo versus the in-vitro experience with coaxial catheter systems is depicted in the figure below. It must be noted that in-vivo friction sometimes needs time to build up - often a couple of minutes - depending on the application and the state of the blood. The blue bars indicate acceptable friction, the yellow bars indicate doubtful performance - strongly depending on duration, vascular curvature, etc. - and the red bars indicate unacceptably high friction.

Some hydrogels dehydrate too quickly when pressure is applied to the coating in an in-vitro test environment. This generally results in a tremendous increase of friction (red part of the bar in the left figure). It is one of the few exceptions in which in-vitro performance is actually worse than the clinical situation.

It is not hard to understand that the deviations in the given example are caused by the differences between the ("living") blood and simple saline (or water). E.g. the low viscosity of the saline often already explains the odd behavior of the hydrogel, i.e. too rapid dehydration. Increase of viscosity to that of blood might improve the in-vitro model significantly. Another effect is found in reactions of the blood with the medical device, often resulting in fibrin-related decay of friction properties.

Sometimes, blood vessels are simulated with help of silicone tubing. Interaction between endovascular catheters and the silicone surface is not by a bit comparable to the interaction with living tissue of the blood vessel. Surface conditions and compliance, as well as the biochemical conditions are completely different. So, one has to address these differences one by one in order to obtain insight in how to translate between in-vitro measurements and in-vivo performance. Many a researcher still fails to see why the "anatomically correct" in-vitro model does not predict clinical performance well enough.


Summarized:
In-vitro modeling must address all in-vivo parameters that are relevant to the functioning of the medical device in order to be successfully predictive. This includes both physical as well as (bio)chemical parameters. The latter is usually wrongfully omitted in practice.