Janus particle-engineered structural lipiodol droplets for arterial embolization

The study addresses limitations of current embolic materials used in minimally invasive arterial embolization therapies for conditions such as hepatocellular tumors, GI bleeding, and renal tumors. Solid beads (>100 µm) poorly deform and struggle to reach finer vasculature, while liquid agents like lipiodol, although radiopaque and drug-loadable, are unstable in blood and prone to recanalization, non-specific embolization, and toxicity. The authors hypothesize that amphiphilic Janus particles, which strongly adsorb at oil–water interfaces, can stabilize lipiodol into mechanically robust, viscoelastic droplets. These engineered droplets should pack efficiently to occlude feeding arteries and deform to traverse and embolize finer arteries, improving efficacy and safety of embolization. The purpose is to design, fabricate, and evaluate Janus particle-engineered structural lipiodol droplets and test their embolization performance in vitro and in rabbit renal embolization in vivo.

Prior work has established embolization as a first-line therapy for various diseases, with both solid microspheres and lipiodol-based liquid embolics widely used. However, complex vascular architectures and size heterogeneity limit distal embolization with rigid beads, while lipiodol emulsions suffer instability in blood, recanalization, off-target embolization, and toxicity. Amphiphilic Janus particles have superior interfacial activity compared to isotropic particles or molecular surfactants and efficiently stabilize emulsions due to high desorption energy and amphiphilicity. Previous reports from the authors demonstrated scalable synthesis of amphiphilic Janus particles with tunable topology and surface chemistry and their ability to stabilize interfaces. Building on these advances, the study leverages Janus particle self-assembly at the lipiodol–water interface to create stable, viscoelastic lipiodol droplets for improved embolization performance.

• Janus particle synthesis: Amphiphilic Janus particles comprising poly(styrene-co-divinyl benzene) (PSDVB, hydrophobic concave side) and poly(acrylic acid) (PAA, hydrophilic convex side) were fabricated via emulsion interfacial polymerization. PS seeds (1.22 ± 0.12 µm) were swollen with 1-chlorodecane, then polymerized with styrene/divinylbenzene/acrylic acid (AIBN initiator, PVA stabilizer) at 70 °C for 14 h. Particles were washed and freeze-dried. SEM: crescent-shaped, 2.26 ± 0.15 µm; zeta potential ~ -42.5 mV.
• Structural lipiodol droplet fabrication: 0.1 mL lipiodol mixed with 6 mL aqueous Janus particle suspension (1.5 mg/mL), sheared 800 rpm for 30 min at room temperature. Janus particles self-assembled at the lipiodol–water interface, yielding droplets (120 ± 40 µm). Size tunability from ~25 ± 16 to 480 ± 53 µm achieved by adjusting particle concentration and oil/water ratios. Scale-up: 200× batch produced ~25 g droplets.
• Characterization: Optical/confocal microscopy (lipiodol labeled with Nile red; Janus particles with amino fluorescein) confirmed dense interfacial coverage. CT imaging assessed radiopacity. Rheology (MCR 302) measured viscosity vs shear rate and oscillatory strain sweeps for storage/loss moduli at 37 °C. Injectability tested with mechanical tester through 2.4 F, 80 cm catheter at 1 mL/min; breakloose/injection forces recorded. Hemocompatibility via hemolysis assay (rabbit RBCs). Cytotoxicity (MTT) with HUVECs; drug-loaded droplet cytotoxicity with HepG2 cells. Drug loading: cisplatin (3–20 mg) mixed into lipiodol prior to droplet formation; loading quantified by ICP-OES; pH-responsive release profiled. Stability assessed over 12 months (size monitoring). Packing capacity: compared with clinical 8spheres beads (100–300 µm) in semi-closed capillary tubes; packing density estimated from packing height.
• In vitro embolization model: Decellularized rat liver prepared by perfusion (H2O, Triton X-100, SDS) until transparent; 1 mL of droplets injected via portal vein (40 µL/s). Distribution and packing in vascular tree assessed by microscopy and CT.
• In vitro capillary flow: Droplet passage and deformation through finer glass capillaries observed under optical microscopy; embolization modeled by injecting droplets at 0.5 mL/min into a sealed-end capillary to study packing and shape evolution under pressure.
• In vivo rabbit renal embolization: New Zealand white rabbits randomized into three groups (n=3/group): Janus droplet group (1.5 mL; ~6000 droplets containing 1 mL lipiodol), 8spheres beads (~30,000 beads, 100–300 µm, 1.5 mL), lipiodol-based emulsion (1.5 mL; O/W 2:1, containing 1 mL lipiodol). Under DSA guidance via femoral access with 2.2 F microcatheter, materials were injected starting from distal renal artery (0.25 mL/s). Iohexol (3 mL) used for pre- and post-embolization DSA on day 0 and day 14. CT arteriography performed for embolization assessment and 3D reconstruction; organ CT screening for off-target embolization. Kidneys harvested at day 14 for gross evaluation.
• Pathology and imaging: Paraffin and frozen sections (4 µm) stained with H&E to assess infarction, vessel occlusion, and droplet deformation within vessels; confocal and SEM to visualize Janus particles in situ. Quantification of embolized vessel diameters; minimum embolized diameter determined.
• Safety and biodistribution: Routine blood, liver/kidney function, coagulation, weight, and inflammatory cytokines (IL-1β, IL-6, PCT, TNF-α) measured pre- and post-embolization (days 0, 3, 9, 14). CT monitoring of lipiodol signal in organs at days 0, 14, 30, 45 to assess lipiodol metabolism and distribution. Janus particle fate studied by Cy7-labeled particles in droplets; fluorescence imaging (IVIS) of organs at day 45; urine examination for particle excretion.
• Microfluidic uniformity (exploratory): Fabrication of uniform droplets (127 ± 7 µm) using microfluidic extrusion of lipiodol into a continuous phase containing Janus particles to self-assemble at the interface.

• Fabrication and properties:
  • Amphiphilic Janus particles (PSDVB/PAA) self-assembled at the lipiodol–water interface to form stable, radiopaque structural lipiodol droplets (mean size 120 ± 40 µm). 1 mL lipiodol produced ~6000 droplets.
  • Size tunable from ~25 ± 16 µm to 480 ± 53 µm by adjusting formulation; long-term stability maintained for at least 12 months (no significant size change).
  • Excellent radiopacity demonstrated by CT; clinical 8spheres beads lacked intrinsic radiopacity.
  • Hemocompatibility: hemolysis rate <2.5% across sizes (<5% considered permissible). Low cytotoxicity to HUVECs up to 32 µg/mL.
  • Drug loading: cisplatin encapsulation efficiency >90% at 6 mg and 10 mg loading; pH-responsive release attributed to PAA carboxyl groups. Cisplatin-loaded droplets increased cytotoxicity to HepG2 cells.
• Viscoelasticity and mechanics:
  • Rheology: strong shear-thinning behavior, facilitating catheter delivery.
  • Storage modulus G' max ~710 Pa for structural lipiodol droplets vs ~3900 Pa for 8spheres beads; lipiodol behaved as liquid. Droplets exhibited higher viscoelastic deformability than beads.
  • Injectability: breakloose and injection forces <10 N for all tested droplet sizes through 2.4 F, 80 cm catheter.
  • Shape adaptability: droplets deformed under stress into ellipsoid, dumbbell, and snowman shapes; passed through finer capillaries via viscoelastic deformation.
  • Packing: in capillary tubes with equal counts and sizes, structural droplets packed to lower height (1.3 cm) vs beads (2.2 cm), corresponding to higher packing density (~0.65 vs ~0.63), implying more efficient vascular occlusion.
• In vitro embolization (decellularized liver): droplets distributed throughout vascular tree, closely packed vessels from feeding arteries to finer branches, demonstrating distal embolization capability.
• In vivo rabbit renal embolization:
  • Delivery: real-time fluoroscopy showed smooth distribution from renal artery into branches; lipiodol emulsion leaked into renal vein and lung (risk of pulmonary embolization). 8spheres required contrast agent for visualization.
  • Efficacy: DSA on day 0 showed near-complete vessel disappearance for droplets and beads. At day 14, droplets maintained efficient embolization (especially renal cortex), whereas lipiodol emulsion showed evident recanalization; 8spheres showed partial recanalization near renal hilum.
  • CT: droplets remained visible in renal cortex at day 14; lipiodol emulsion largely metabolized with low CT signal. 3D reconstructions confirmed persistent embolization.
  • Kidney volume reduction rates (day 14, CT volumetry): structural droplets 38.85%, 8spheres 38.56%, lipiodol emulsion 8.06% (n=3; ***p<0.001 vs control where indicated; ns p=0.16 for certain comparisons).
  • Gross pathology: embolized right kidneys were milky white and reduced in size for droplet and bead groups; lipiodol emulsion kidneys remained largely blood-red with limited reduction.
  • Histology: droplet group showed extensive cortical necrosis consistent with ischemic infarction; 8spheres group had hyperemic zones indicating recanalization and some normal glomeruli; emulsion group showed largely normal structures. Droplets were closely packed in vessels including fine vasculature; deformation evident in situ. Minimum embolized vessel diameter ~40 µm. Confocal/SEM confirmed abundant Janus particles within vascular lumens; no micro-emboli detected. 8spheres primarily lodged in larger arteries, not finer vessels.
  • Safety/off-target: CT of other organs at day 14 showed no non-target embolization (lung, liver, heart, spleen, brain, left kidney) in droplet group. Routine blood, liver/kidney function, coagulation, weight, and inflammatory cytokines remained within normal ranges post-embolization.
• Fate and metabolism:
  • Lipiodol signal decreased over time (days 0–45), indicating gradual metabolism; CT showed signal localized to kidney and bladder over time.
  • Cy7-labeled Janus particles detected in kidney, liver, and bladder at day 45; particles also present in urine, suggesting clearance pathways via liver and kidney.
• Manufacturability: approach scalable to produce ~25 g per batch; droplets easily packaged. Microfluidic method yielded uniform droplets (127 ± 7 µm).

The findings validate the hypothesis that amphiphilic Janus particles can stabilize lipiodol into mechanically robust, viscoelastic droplets capable of efficient arterial embolization. Dense particle shells confer structural integrity for tight packing and sustained occlusion of feeding arteries, while viscoelastic deformation allows navigation into and embolization of finer vasculature down to ~40 µm, overcoming limitations of rigid beads. Compared with clinical lipiodol-based emulsions, the engineered droplets demonstrated superior in vivo persistence without recanalization, intrinsic radiopacity for real-time guidance, and reduced risk of non-target embolization. The comparable kidney volume reduction to 8spheres, coupled with better distal embolization and visualization, underscores clinical relevance. Safety assessments indicated good hemocompatibility and absence of systemic toxicity or off-target emboli at 14 days. The demonstrated high drug-loading efficiency and pH-responsive release further suggest utility for chemoembolization. Overall, the results support the use of Janus particle-engineered structural lipiodol droplets as a versatile embolic platform addressing key shortcomings of current materials.

This work presents a scalable strategy to fabricate radiopaque, viscoelastic structural lipiodol droplets by programming amphiphilic Janus particle self-assembly at the lipiodol–water interface. The droplets closely pack to occlude feeding arteries and deform to embolize finer vessels, achieving efficient renal embolization in rabbits with no recanalization or non-target embolization over 14 days. They show favorable injectability, hemocompatibility, long-term stability, high drug-loading with pH-responsive release, and real-time imaging visibility. The platform holds promise for clinical embolization, particularly TACE, and can be combined with therapeutics. Future directions include standardizing droplet size distribution via microfluidics (demonstrated 127 ± 7 µm), expanding to other vascular beds and disease models, longer-term safety/effectiveness studies, and optimizing drug delivery profiles.

• Size distribution: The bulk-fabricated droplets had a relatively wide size distribution (120 ± 40 µm), potentially complicating particle selection and predicting embolic efficacy; a microfluidic approach was proposed to achieve uniform sizes (127 ± 7 µm).
• Biodegradability: The Janus particles are non-degradable; although metabolic/biodistribution studies suggest clearance via liver and kidney with urinary excretion, long-term fate beyond 45 days and potential accumulation require further evaluation.
• Study scope: In vivo validation was limited to rabbit renal embolization with small group sizes (n=3 per group) and 14-day primary efficacy readout; broader applications, longer-term outcomes, and comparative dosing regimens were not assessed.
• Comparative imaging: 8spheres beads lack intrinsic radiopacity, necessitating contrast for DSA; differences in imaging modalities may influence procedural assessment.