A stent that does its job, then gradually gets out of the way entirely
A bioresorbable vascular scaffold (BVS) is deployed like any balloon-expandable stent — but its entire structural body, not just a drug-eluting coating, is built from a resorbable polymer. At the moment of implantation it provides radial mechanical support comparable to a permanent metal stent, holding the artery open against elastic and plastic recoil. The difference only becomes apparent over the following months, as the device is designed to do its job and then progressively disappear.
Permanent metal stents solved acute recoil and, with drug-eluting technology, greatly reduced restenosis — but they leave a rigid metal cage in the artery forever. That permanent cage has real costs: it fixes the vessel's diameter, preventing the natural vasomotion (dilation and constriction in response to exercise, nitrates, or endothelium-derived signals) that a healthy artery uses to match blood flow to demand. It also creates a permanent nidus for late complications — very late stent thrombosis, neoatherosclerosis within the stented segment, and chronic low-grade inflammation around the durable polymer coating — and can complicate future imaging (CT artifact) or surgical revascularization at that site.
The idea behind a fully bioresorbable scaffold — sometimes marketed as "vascular restoration therapy" — is to invert this trade-off: provide full mechanical scaffolding only for as long as the vessel actually needs it during acute healing, then let the device disappear entirely, handing mechanical responsibility back to a remodeled, healed artery wall. In principle this should combine the acute benefits of stenting with a long-term result indistinguishable from an artery that was never stented at all.
The appeal is straightforward: a stent that works for six to twelve months and then no longer exists is, in theory, better than a stent that works for six to twelve months and then exists for the rest of the patient's life.
It is important to distinguish a BVS from a conventional biodegradable-polymer drug-eluting stent (DES). In a biodegradable-polymer DES, only the thin polymer coating that carries and releases the drug resorbs — the metal strut backbone (cobalt-chromium or platinum-chromium) remains permanently in the vessel wall. In a BVS, the entire strut backbone itself is the resorbable polymer.
Absorb BVS was manufactured by extruding PLLA into a tube, laser-cutting it into the same open-cell strut pattern used on metal stents, then coating it with a thin layer of poly-D,L-lactic acid (PDLLA) loaded with everolimus — the identical antiproliferative drug used on Abbott's metallic Xience stent, released with similar kinetics (roughly 80% eluted within the first 30 days). This design let manufacturers isolate the variable of interest — a fully resorbable backbone — while keeping the antiproliferative strategy unchanged from a proven metal platform.
During the first six months, the scaffold's job is essentially identical to that of a permanent metal stent: resist acute and sub-acute vessel recoil while the injured arterial wall heals. Hydrolysis of the polymer backbone begins immediately upon implantation, but early on it manifests mainly as a fall in molecular weight — the mechanical, load-bearing properties of the strut lag well behind.
PLLA degrades primarily through bulk hydrolysis: water diffuses into the amorphous regions of the polymer and cleaves the ester bonds along the backbone, progressively shortening the polymer chains. Molecular weight therefore begins declining from day one.
Mechanically, though, the strut behaves like a coherent solid for much longer than the molecular weight curve alone would suggest. A polymer's bulk strength depends on chain entanglement — long chains physically interlocking to resist deformation — and this network stays largely intact until enough chain scission events accumulate to push average chain length below a critical entanglement threshold. Below that threshold, strength falls away rapidly. Above it, radial strength is buffered even as the underlying molecular weight is already falling substantially.
The first weeks after implantation carry real elastic and plastic recoil risk, as the balloon-injured vessel wall responds to the trauma of deployment. The scaffold must resist this recoil while endothelial and neointimal coverage of the struts proceeds — smooth muscle cell migration and extracellular matrix deposition gradually embed each strut beneath a layer of living tissue, typically visualized serially with optical coherence tomography (OCT), which by 6–12 months shows the great majority of struts fully covered.
This is also the period in which "late lumen loss" — a modest reduction in lumen diameter from neointimal growth over the struts — is most actively occurring, making continued mechanical support from the scaffold clinically important even as its polymer chemistry is already changing underneath.
Between roughly six and twelve months, the gap between molecular-level degradation and mechanical failure finally catches up: radial strength falls sharply as molecular weight drops below the entanglement threshold across most of the strut cross-section. Mechanical load is transferred, month by month, from the weakening scaffold to the arterial wall it has been holding open.
This is the single most important distinction in bioresorbable scaffold degradation: loss of radial strength (mechanical failure) and loss of mass (full resorption) are not the same process, and they do not happen on the same timescale.
Radial strength depends on chain entanglement, which fails once molecular weight crosses a critical threshold — a comparatively fast, nonlinear transition. Full mass loss requires the polymer to be broken down all the way to soluble oligomers and ultimately lactic acid monomers small enough to diffuse out of the strut and be cleared — a much slower, diffusion-limited process. The practical consequence is a prolonged window, roughly spanning months 6 through 24, during which the scaffold has largely stopped providing meaningful mechanical support but is still physically present as bulk polymer occupying space in the arterial wall.
This mismatch — mechanically absent, physically present — is now understood to be the central engineering challenge of bioresorbable scaffold design, and a major contributor to the adverse events later seen in first-generation clinical trials.
A scaffold can be a mechanical non-entity while still being a physical, and potentially thrombogenic, obstacle inside the vessel — this gap between strength loss and mass loss defines the highest-risk period of the device's life cycle.
While the scaffold is losing its mechanical role, the artery is — ideally — gaining one. Positive vascular remodeling continues: a mature, organized neointima made of smooth muscle cells and collagen-rich extracellular matrix consolidates over and around the struts, the internal and external elastic laminae adapt to the new lumen geometry, and by roughly twelve months serial intravascular ultrasound (IVUS) and OCT studies generally show a vessel wall capable of bearing pulsatile hemodynamic load largely on its own — precisely the state the entire therapy is designed to reach before the scaffold's support becomes unreliable.
With radial strength already largely gone, the scaffold ring itself begins to lose circumferential continuity — struts fracture into discrete fragments embedded in the arterial wall, which then undergo slow surface erosion as hydrolysis continues. This is also the period in which the largest randomized trials of first-generation bioresorbable scaffolds revealed a troubling safety signal.
As hydrolysis proceeds through the remaining amorphous polymer, the once-continuous circumferential ring of struts loses mechanical continuity and fractures into discrete fragments. These fragments become embedded within the surrounding neointima and media, where they behave as inert particulate matter undergoing progressive surface erosion — shrinking gradually as their outer layers hydrolyze to soluble lactic acid, which is cleared systemically and metabolized through the Krebs cycle to carbon dioxide and water, the same pathway the body uses for endogenous lactate.
The ABSORB clinical program (ABSORB II, III, and IV, enrolling thousands of patients against the metallic everolimus-eluting Xience stent as the active comparator) was designed to demonstrate non-inferiority of the fully bioresorbable Absorb BVS. Instead, pooled and individual trial results showed a consistent, statistically meaningful excess of target lesion failure and — more concerning — device (scaffold) thrombosis in the Absorb arm, including "very late scaffold thrombosis" occurring more than a year after implantation, an event rarely seen with modern metallic drug-eluting stents.
The leading mechanistic explanations converge on exactly the strength/mass mismatch described in Stage 3: Absorb's struts, at roughly 150 µm, were nearly twice as thick as contemporary metal DES struts (~80–100 µm), disturbing blood flow and delaying full endothelial coverage; and the prolonged fragmentation phase left discontinuous, sometimes malapposed strut remnants sitting in the bloodstream as a thrombogenic surface for many months after they had ceased contributing any mechanical benefit. Undersizing in small vessels (<2.25 mm) compounded the problem, amplifying flow disturbance around the relatively bulky struts.
These findings led Abbott to voluntarily withdraw Absorb BVS from worldwide commercial sale in September 2017, just over a year after its 2016 FDA approval — a stark lesson that a sound therapeutic concept can still fail on execution details of strut geometry and degradation timing.
By roughly 24–36 months, the polymer is, ideally, fully resorbed — no permanent implant remains in the artery. The theoretical prize of this entire technology class is a vessel that behaves as though it was never stented: free to dilate and constrict in response to physiological stimuli, and free of the chronic foreign-body burden that comes with a lifelong metal cage. First-generation clinical experience showed this prize is real but only partially, and inconsistently, realized.
Longer-term substudies of the Absorb program (including serial OCT/IVUS imaging and pharmacologic vasomotion testing with agents such as acetylcholine and nitroglycerin) did demonstrate, in a subset of patients followed to three to five years, an absence of persistent visible scaffold material and some restoration of vasomotor responsiveness compared with a rigid, permanently caged segment. This is the mechanistic proof-of-concept the entire "vascular restoration therapy" idea rests on.
However, this long-term benefit was not sufficient to offset the excess early-to-mid-term thrombotic and target lesion failure risk observed in the larger randomized trials — a reminder that a device must be judged across its entire life cycle, not only at its intended endpoint.
The first-generation BVS experience reframed the engineering problem for the whole field: it is not enough to make a scaffold fully resorbable — the timing relationship between strength loss, fragmentation, and mass loss has to be compressed and better matched to strut geometry, so the device never spends many months mechanically absent yet still physically obstructive.
Subsequent designs have pursued this directly: thinner-strut PLLA-based platforms (approaching ~100 µm, closer to modern metal DES), alternative backbone chemistries such as tyrosine-derived polycarbonate (radiopaque, thinner-strut designs), and magnesium-alloy resorbable scaffolds (which corrode via a different, faster electrochemical mechanism, largely resorbing within about a year and avoiding a prolonged polymer-fragment phase altogether). Across all of these, the explicit design target is the same: shrink or eliminate the vulnerable window between "no longer mechanically useful" and "not yet gone."
The underlying biology of the vulnerable window — molecular weight loss, then strength loss, then mass loss, each on its own timescale — has not changed. What next-generation devices are trying to change is the engineering: thinner struts and better-tuned degradation kinetics so the mismatch identified in first-generation BVS technology no longer translates into excess clinical risk.