The study, published in Frontiers in Biomaterials Science, centers on a modified form of dextran, a sugar-based polymer with a long clinical history as a plasma expander and drug carrier. Dextran is built from glucose units linked predominantly through alpha-1,6 bonds, and each unit carries three hydroxyl groups that chemists can exploit. The Drexel team attached methacrylate groups to these hydroxyls by reacting dextran with glycidyl methacrylate, a base-driven esterification that proceeds uniformly in solution and allows the degree of substitution—the fraction of repeat units bearing a reactive methacrylate group—to be dialed anywhere from 5 to 95 percent. Nuclear magnetic resonance spectroscopy confirmed that this degree of substitution scaled linearly with the amount of glycidyl methacrylate used, giving the researchers a precise handle over how many light-reactive sites each polymer chain would carry.

The material’s defining trick lies in a two-stage photopolymerization scheme. In the first stage, the dissolved polymer and a water-soluble photoinitiator called lithium phenyl-2,4,6-trimethylbenzoylphosphinate, or LAP, are exposed to blue light (400 to 500 nanometers) just long enough—about 20 seconds—to form a solid, self-supporting hydrogel. Critically, this brief cure is deliberately incomplete: many methacrylate groups remain unreacted, held in reserve inside the swollen network. After the gel equilibrates overnight in salt solution, a second, much longer exposure of 60 to 600 seconds triggers a fresh wave of crosslink formation. Every new covalent bridge increases the network’s crosslink density, which squeezes water out of the gel. The hydrogel contracts, and any tube lined with the coating sees its inner diameter grow.

Rheological measurements traced the physics of this transformation in real time. Dissolved DexMA began as a weakly viscous fluid with a storage modulus below 1 kilopascal; within minutes of illumination, the storage modulus climbed steeply and then plateaued as available vinyl groups were consumed. Higher photoinitiator concentrations (3 and 5 millimolar) and higher light intensities (5 and 10 milliwatts per square centimeter) produced stiffer final gels—but they also exhausted the reactive pool too quickly, leaving nothing for the second act. The team therefore standardized on the mildest conditions, 1 millimolar LAP and 1 milliwatt per square centimeter, for the casting step. Under a 20-second initial cure, the gel reached roughly 1 kilopascal, and interestingly, crosslinking continued mildly after the lamp switched off, a sign that reactive radicals were diffusing through the network and extending growing polymer chains.

The parameter space governing how much the gel can shrink proved rich and, in places, counterintuitive. Volumetric changes rose with polymer concentration, from 2.5 up to 20 percent weight per volume, simply because denser solutions carry more methacrylate groups per unit volume. But the degree of substitution followed a more subtle pattern: gels made with intermediate modifications—25 to 50 percent—showed the largest volume changes and the widest mechanical tunability, while highly substituted polymers performed worse. The likely culprit is hydrophobicity. At high methacrylate densities, the modified dextran chain effectively collapses, reducing its hydrodynamic radius and the entanglements that support network formation, so final moduli actually fell for modifications exceeding 50 percent despite the abundance of reactive vinyl groups. Light dose also mattered temporally: significant differences appeared among 60-, 300-, and 600-second cures, though the response saturated between 300 and 600 seconds, meaning surgeons could dial in a specific degree of contraction simply by controlling exposure time.

Under optimized conditions—a 20-second casting at low initiator followed by a 300-second cure at 5 millimolar LAP—the hydrogels achieved volume reductions of 20.4 percent, plus or minus 0.79 percent. That number is not arbitrary. Earlier computational fluid dynamics work by the same group showed that a 20 percent volume change in a hydrogel lining translates to a 15 to 18 percent increase in shunt lumen diameter, which is precisely what models predict a child needs to accommodate the first two years of growth. A published case study of a dilatable shunt, whose lumen was expanded from 2.7 to 3.3 millimeters—an 18.2 percent expansion to correct inadequate blood perfusion—lends clinical corroboration to those targets. The team demonstrated feasibility in a benchtop model by loading a 50-percent-modified hydrogel into surface-treated quartz tubing with a 3.5-millimeter lumen and pulsing it with light in 10-second increments; the required 15 to 18 percent diameter increase arrived within 50 to 70 seconds of cumulative exposure.

An implantable material must also survive the body, and here the results were reassuring. Over three months of incubation at physiological temperature and pH, every formulation degraded by less than 10 percent cumulatively, swelled or shrank by less than 5 percent in volume, and lost minimal mass—with only the softest, 2.5-percent gels showing moderate losses approaching 25 percent. Neither the dextran molecular weight (75 versus 500 kilodaltons) nor the degree of substitution materially altered this hydrolytic resilience, which the authors attribute to the stability of dextran’s glycosidic bonds and the dense crosslinked network. Because the geometry of the gel remains essentially locked in place for months, a light-induced adjustment would be expected to persist rather than drift back.

Biological testing spanned the cell types an implanted shunt would actually meet. NIH 3T3 fibroblasts proliferated normally for three days in direct contact with 25- and 50-percent-modified gels and in their extracts, with population doubling times matching the roughly 20 hours typically reported, though the 95-percent formulation depressed metabolic activity—likely because hydrophobic, highly substituted soluble polymer leached into the medium at concentrations exceeding anything a pre-polymerized gel could realistically release. RAW-Blue macrophages, a reporter line for innate immune activation, adhered poorly to DexMA surfaces compared with polystyrene and showed minimal inflammatory signaling after 24 hours across all formulations, casting and curing alike. Most importantly for a blood-contacting device, ASTM-compliant hemolysis testing with citrated whole rabbit blood returned hemolytic indices below 3 percent for every gel tested—comparable to the inert low-density polyethylene negative control, far below the Buna-N rubber positive control, and safely under the 5 percent threshold the standard recommends.

The implications reach beyond pediatric cardiology. A hemocompatible gel whose volume can be sculpted after fabrication, on a timescale of seconds to minutes, could serve in dynamically adaptable cell-culture platforms, size-changing implants, or any blood-interface construct where geometry must evolve. The authors are candid about remaining hurdles: LAP is a small, water-soluble molecule whose retention inside a gel under active blood flow cannot be assumed, so clinical implementation would likely require delivering fresh photoinitiator locally through a catheter before irradiation—conveniently, the same blue-light-emitting catheter envisioned for the adjustment itself. Even so, by mapping the full parameter space of polymer concentration, methacrylate density, initiator loading, and light dose, the Drexel team has transformed a clever proof-of-concept into an engineering manual. For the youngest cardiac patients, that could one day mean their implant grows up with them, one carefully metered pulse of light at a time.

Subject of Research: Photoresponsive, hemocompatible dextran hydrogels with tunable volumetric contraction for pediatric blood shunts

Article Title: Hemocompatible polymeric hydrogels with tunable photoresponsive volumetric properties

Article References: Seiner, A. J., Mishin, V., Soni, S. S., Posthill, E., Sarkar, B., Spiller, K. L., Throckmorton, A. L., & Rodell, C. B. (2026). Hemocompatible polymeric hydrogels with tunable photoresponsive volumetric properties. Frontiers in Biomaterials Science, 5, Article 1924191. https://doi.org/10.3389/fbiom.2026.1924191

Image Credits: AI Generated

DOI: 10.3389/fbiom.2026.1924191

Keywords: hydrogels, dextran, photopolymerization, pediatric cardiology, blood shunts, hemocompatibility, biomaterials, congenital heart defects, photoresponsive materials, crosslink density, medical devices, tunable implants