Flow Diversion – Strategy, Complications, and Innovation

This session offered one of the most comprehensive overviews of real-world flow diverter (FD) deployment challenges and evolving adjunctive technologies. The session began with a prospective study analysing 245 FD deployments (Bicetre Experience) and categorizing braid deformations into four primary types: fishmouthing (39%), collapse (25%), shortening (20%), and bumps (16%). Although these deformations didn’t impact overall occlusion rates or direct morbidity, patients with early braid deformations requiring retreatment showed higher delayed morbidity.

A key insight came from Oliver Babeau, design engineer from Balt, who demonstrated in a silicon model how oversized FDs, even if initially well-deployed, are prone to delayed distal fishmouthing and prolapse, especially when deployed in vessels with fusiform dilations. His lab demonstration using a 3.75 mm and 2.25 mm FD in a 2 mm vessel was striking. The oversized FD gradually slipped into the aneurysmal sac due to unbalanced radial forces, while the smaller, appropriately sized FD remained well-positioned.

In my personal practice, I usually try to avoid oversizing, especially when there’s a significant mismatch between proximal and distal landing zone diameters. A large taper is often a red flag prompting me to consider alternatives such as stent-assisted coiling or even microsurgical options. I prioritize selecting an FD that matches the proximal landing zone diameter as closely as possible and avoid compromise at the distal end.

Professor Vincent contributed a helpful rule of thumb that I used to follow extensively while working with single-plane systems: the final deployed length of an appropriately sized FD is approximately one-third the length of the device when inside the microcatheter. Although I now rely more on the simulator embedded in our Artis Zee biplane system, I still find this rule invaluable when visual guidance is limited or uncertain.

The discussion introduced a critical classification of fishmouthing into acute, subacute, and chronic forms. Acute types often manifest intraoperatively and are mechanical in origin, correctable with manipulation or balloon angioplasty. Subacute types appear in the hours or days post-deployment and can lead to vessel occlusion. Chronic fishmouthing, however, stood out. The panel shared that it is more commonly seen in young females and in the supraclinoid ICA region, possibly due to vascular hyperreactivity, intimal hyperplasia, or chronic muscle spasm.

This resonated deeply with a case I personally treated, a ruptured supraclinoid ICA blister aneurysm in a middle-aged woman. At 6-month follow-up, angiography revealed significant narrowing at the distal end of the FD, which at the time puzzled me. After this session, the picture became clearer, it likely reflected chronic fishmouthing due to reactive remodelling. Such post-hoc realizations help shape future expectations and follow-up protocols.

Professor Spelle demonstrated a clever workaround in cases with short distal landing zones near bifurcations: deploying a laser-cut stent beyond the distal FD end to reinforce anchorage. This was shown with impeccable precision in a live model and is a strategy I will consider for similarly constrained anatomies.

The introduction of the neuro-specific OCT probe (nOCT) added an exciting layer to the session. Compatible with a 0.021” microcatheter, it allows intra-procedural imaging of:

• Wall thickness to assess rupture risk zones

• Stent apposition verification immediately post-deployment

• Early thrombus formation detection

• Neointimal growth and stent response during follow-ups

While we do not currently use nOCT in our setup, I see clear value in adding it to our armamentarium in the future, particularly for high-risk fusiform aneurysms and when treating under-covered segments.

Dr. Adnan Siddiqui also emphasized the role of vessel wall imaging in evaluating proximal stenosis adjacent to large aneurysms. The imaging can help differentiate between atherosclerotic narrowing (which shows thickened walls and enhancement) versus mechanical kinking (which may appear normal). In my current practice, I often rely on forward microcatheter pressure during stent deployment to force open stenosed segments, a workaround that has yielded consistent results but would be better supported by more robust pre-op imaging.

This session was not just about technicalities; it was about evolution in understanding, learning to recognize subtle complications, and integrating imaging as a central component of planning. It reminded me that flow diversion is not simply a device deployment, it’s a dynamic interaction between stent, vessel wall, and hemodynamics.