Abstract
Background/Aim: Poly lactic acid microsphere fillers stimulate neocollagenesis, but aggregation can yield heterogeneous remodeling. We compared poly-L-lactic acid (PLLA)- and poly-D,L-lactic acid (PDLLA)-based fillers to connect stereochemistry with microsphere stability, dispersion, and in vivo outcomes.
Materials and Methods: Poly lactic acid/sodium hyaluronate composite fillers were analyzed by scanning electron microscopy, nuclear magnetic resonance, differential scanning calorimetry, and micro-compression. Dispersion-aggregation-redispersion in saline was quantified over 3 days by optical microscopy with circularity-based analysis. In SKH mice, 100 μl filler was injected subcutaneously. Projection volume was measured at baseline (day 0) and at 2, 4, 8, and 12 weeks post-injection using phase-shift rapid in vivo measurement of skin (PRIMOS). Tissue response was observed at the same time points using hematoxylin and eosin, Masson’s trichrome, and macrophage immunofluorescence.
Results: PLLA microspheres remained discrete after saline dispersion and excipient removal, whereas PDLLA particles deformed and collapsed into compact clusters. PLLA particles (median 49.10 μm) were larger than PDLLA particles (median=19.85 μm) and showed semicrystalline differential scanning calorimetry features (glass transition temperature=60.88°C, cold crystallization temperature=94.2°C, melting temperature=169.18°C), while PDLLA was predominantly amorphous. PDLLA formed larger, poorly re-dispersible aggregates (median cluster=458.06 μm) than PLLA (median cluster=380.02 μm). In vivo, volume declined through week 4 then recovered. PLLA rebounded more uniformly with greater collagen area and weaker inflammatory and macrophage signals than PDLLA.
Conclusion: PLLA crystallinity and mechanical robustness support re-dispersibility and more homogeneous collagen remodeling, whereas PDLLA aggregation is linked to heightened inflammation and reduced collagen deposition. Collectively, these findings suggest that maintaining microsphere integrity and dispersion is a key, actionable determinant of more uniform biostimulatory outcomes in PLA-based fillers.
Introduction
Minimally invasive injectable fillers are widely used to address loss of soft-tissue volume and to improve skin quality (1-3). Among filler classes, hyaluronic acid (HA) fillers are valued for their immediate volumizing effect and the possibility of enzymatic reversal such as via hyaluronidase-assisted degradation (4, 5), whereas biostimulatory fillers are designed to induce long-term volume improvement by promoting neocollagenesis and extracellular matrix remodeling over time (6). Poly lactic acid (PLA) is one of the most established collagen-stimulating materials in esthetic and plastic surgery and has been extensively investigated across facial and off-facial indications (7, 8).
PLA microspheres trigger a controlled foreign-body response that recruits macrophages and fibroblasts to deposit collagen as the polymer degrades (9). Clinical studies have supported the efficacy of injectable poly-L-lactic acid (PLLA) for correction of facial folds compared with HA-based fillers, demonstrating meaningful aesthetic improvement with an extended effect profile when appropriately prepared and injected (10, 11). More recently, poly-D,L-lactic acid (PDLLA) microsphere fillers have gained attention as biostimulatory injectables [reviewed in (12)], and randomized, evaluator-blinded multicenter clinical evidence has continued to accumulate for PDLLA-based products in the treatment of nasolabial folds (13-15).
Despite their demonstrated benefits, collagen-stimulating fillers present practical and safety challenges that are closely linked to particle distribution in tissue [reviewed in (16)] and supported by experimental reports (17). Unlike HA gels, PLA particles cannot be readily reversed by enzymatic action, such as hyaluronidase, and therefore maldistribution or particle aggregation may persist and contribute to the development of palpable nodules or papules (18). For both PLLA and PDLLA microsphere systems, expert consensus and clinical guidance emphasize that reconstitution, dilution and mixing technique are not trivial handling steps but key determinants of injectability, even spread, and mitigation of adverse events (19). These observations suggest that suspension stability and particle re-dispersibility are key intermediates linking material design and handling to clinical outcomes.
From a biomaterials perspective, an underexplored factor in injectable PLA microsphere performance is polymer stereochemistry, which governs crystallinity and thereby influences thermal behavior, mechanical properties, water transport, and degradation kinetics [reviewed in (20)]. The difference between these fillers is that PLLA is semicrystalline, whereas PDLLA is largely amorphous (21). Crystalline PLA generally has higher thermal transitions and mechanical rigidity than amorphous PLA, affecting particle stiffness and water uptake (22). We hypothesized that these differences would cause PLLA and PDLLA microparticles to behave differently in aqueous environments, shaping implant microarchitecture and remodeling. To test this, we compared PLLA/HA composite fillers and PDLLA/HA composite fillers. We examined (i) particle morphology by scanning electron microscopy (SEM), (ii) polymer stereochemistry by proton nuclear magnetic resonance (1H NMR), (iii) thermal transitions by differential scanning calorimetry (DSC), and (iv) mechanical deformation by micro-compression. We quantified particle dispersion and aggregation in saline over time using optical microscopy and circularity analysis. Also, we injected each filler subcutaneously in mice and assessed volumetric retention and tissue response. By linking polymer stereochemistry and particle properties to suspension behavior and tissue remodeling, this work aims to explain how PLA fillers with similar composition can diverge in clinical performance.
Materials and Methods
Materials. PLLA particles (InCube 701®, a commercialized product), and PDLLA particles especially prepared for this comparative study were supplied by by LabInCube Co., Ltd. (Seoul, Republic of Korea). Sodium hyaluronate was purchased from SHANDONG AWA BIOPHARM Co., Ltd. (Shandong, PR China).
Sample preparation. PLLA and PDLLA fillers were prepared by mixing PLA particles with sodium hyaluronate in a phosphate-buffered aqueous solution (PBS; pH 7.0, adjusted with monobasic and dibasic sodium phosphate). The suspension was formulated to contain 85.0% (w/w) PLLA particles and 12.4% (w/w) HA, gently stirred until homogeneous, and then frozen to dry.
Separation of PLA particles from aqueous excipients. To isolate PLA particles from the aqueous excipient phase containing HA, PLLA or PDLLA fillers (200 mg) were suspended in 10 ml of distilled water and vortexed until a homogeneous suspension was obtained. The suspension was centrifuged at 6,300×g for 5 min, the supernatant containing water-soluble excipients was discarded, and an equal volume of distilled water was added; this washing step was repeated three times. To further remove residual moisture, an equal volume of ethanol was added, and the same centrifugation and decanting procedure was performed. The washed PLA particles were finally dried under vacuum at room temperature and collected as a dry powder for subsequent analyses.
SEM. The morphology of the dermal fillers and the size of particles were examined using a field-emission scanning electron microscope (JSM-7600F; JEOL Ltd, Tokyo, Japan) at the Chronic and Metabolic Diseases Research Center, Sookmyung Women’s University. For SEM observation, three sample preparation methods were used: (i) filler samples were directly mounted onto copper tape on aluminum stubs and sputter-coated with platinum; (ii) filler samples were dispersed in saline, drop-cast onto silicon wafers, dried under vacuum, then sputter-coated with platinum; and (iii) particles isolated from aqueous excipients were dispersed in ethanol, drop-cast onto silicon wafers, dried under vacuum, then sputter-coated with platinum before imaging.
For particle-size distribution analysis, at least 200 individual microspheres were randomly selected from SEM micrographs, and their diameters were measured using image-analysis software (ImageJ, National Institutes of Health, Bethesda, MD, USA). The resulting particle size distributions were fitted to a log-normal distribution model.
NMR. The monomer configuration of PLA particles was characterized by 1H NMR spectroscopy to distinguish between PLLA and PDLLA. Spectra were recorded on an Avance III HD 500 NMR spectrometer (500 MHz; Bruker BioSpin, Bruker, Rheinstetten, Germany) at the Chronic and Metabolic Diseases Research Center, Sookmyung Women’s University, using deuterated chloroform (CDCl3) as the solvent. Samples were prepared by dissolving 10 mg of PLA particles in 1 mL of CDCl3, and the resulting 1H NMR spectra were qualitatively analyzed to determine the stereochemical configuration of the LA units.
DSC. The thermal properties and crystallinity of PLA particles were analyzed by DSC to compare semi-crystalline PLLA with amorphous PDLLA. Samples of PLLA particles separated from PLLA fillers (2.7350 mg) and PDLLA particles derived from PDLLA fillers (2.4430 mg) were sealed in aluminum pans and measured using a DSC Q20 instrument (TA Instruments, Newcastle, DE, USA) at the Korea Polymer Testing and Research Institute. The samples were heated from 0 to 210°C at a rate of 10°C/min under a nitrogen atmosphere, and the resulting DSC thermograms were used to evaluate glass transition, cold crystallization, and melting behavior of the PLA particles.
Micro-compression testing. The compressive properties of PLA dermal filler particles were evaluated at the Korea National University of Transportation, Central Laboratory using a micro-compression testing machine (MCT-510; Shimadzu, Kyoto, Japan) equipped with a flat indenter (FLAT50) and a 50× objective lens. For PLLA fillers and particles isolated from filler, individual particles were compressed at a loading speed of 10 gf/s up to the breaking point, with a maximum test force of 30 gf and 5-6 particles were measured for each condition. For PDLLA fillers and particles, a loading speed of 5 gf/s was applied, with maximum test forces of 10 gf for the fillers and 1 gf for the particles and 5-6 particles were tested for each condition. The compression ratio was set to 10 for all tests, and load–displacement data were continuously recorded and converted to pressure–strain curves.
Dispersity analysis. PLA filler samples were prepared by suspending 170 mg of PLA in 4, 8, 12, or 16 ml of saline. The suspensions were stored at room temperature for up to 3 days. At each time point (0, 1, 2, and 3 days), an aliquot of each suspension was placed on a glass slide, the observation area was divided into a 3×3 grid, and one optical microscopy (OM) image was acquired from each of the nine predefined regions. On days 1-3, the suspensions were first gently hand-shaken and imaged by OM and were then vortexed for 3 min followed by a second OM imaging.
For each time point and each region, particle morphology was analyzed from OM images using ImageJ (National Institutes of Health). The circularity of individual particles was quantified, and objects with circularity values below a predetermined threshold were classified as clusters. For each sample, the degree of dispersion was calculated as the fraction of well-dispersed particles (non-clustered particles) among all particles, and the cluster size distribution was obtained from the measured sizes of clusters.
Animal study. All animal procedures were approved by the Institutional Animal Care and Use Committee of Seoul National University Bundang Hospital (approval no. BA-2502-409-004-01) and were performed in accordance with institutional and national guidelines. Five-week-old male SKH mice (20-25 g; BioOrient, Seongnam, Republic of Korea) were housed in a specific pathogen-free facility under a 12 h light/dark cycle at 24°C and 55% relative humidity, with ad libitum access to food and water.
A total of 75 mice were randomly assigned to three groups: PBS, PDLLA, and PLLA (n=25 per group). Each mouse received a subcutaneous injection of 100 μl of the assigned material into the dorsal skin. Primos images (Canfield Scientific, Parsippany, NJ, USA) were acquired for all animals at baseline (uninjected flat area) and at 2, 4, 8, and 12 weeks. For histological evaluation, five mice per group were euthanized at each time point. Consequently, the numbers of animals contributing to volume measurements per group were 25, 20, 15, 10 and 5 at 0, 2, 4, 8, and 12 weeks, respectively. Then a full-thickness skin and subcutaneous tissue block encompassing the injection site (centered at the injection point; approximately 0.3 cm×0.3 cm) was harvested, fixed in 10% neutral buffered formalin, and processed for paraffin embedding.
Hematoxylin and eosin staining. Paraffin-embedded sections collected at the above time points were deparaffinized in xylene and rehydrated through a graded ethanol series (100%, 95%, 90%, 80%, and 70%). After rinsing in distilled water, sections were stained with hematoxylin for 5 min, followed by bluing for 10-15 s. Eosin was then applied for 30 s to stain cytoplasmic and extracellular components. Sections were subsequently dehydrated in 95% and 100% ethanol, cleared in xylene, and mounted with coverslips. Images were acquired using an Olympus light microscope (Olympus Corporation, Tokyo, Japan). Neovascularization and inflammatory cell infiltration were evaluated semi-quantitatively by ImageJ (National Institutes of Health). Neovascularization was quantified within a predefined region of interest (ROI) encompassing the implant area and a 3 mm peri-implant margin by selecting three non-overlapping high-power fields per section and calculating the red blood cell area fraction using ImageJ. Inflammatory cell infiltration was assessed within the same ROI by counting inflammatory cells in three high-power fields (400×) per section.
Masson’s trichrome staining. Masson’s trichrome staining was performed to evaluate collagen deposition. Paraffin sections were incubated in Bouin’s solution overnight at room temperature, rinsed, and then stained with Weigert’s hematoxylin for 10 min. Sections were then stained with Biebrich scarlet-acid fuchsin to label cytoplasm and muscle fibers, followed by differentiation in phosphomolybdic-phosphotungstic acid. Collagen fibers were counterstained with aniline blue, briefly rinsed in water, treated with 1% acetic acid, and subsequently dehydrated, cleared in xylene, and mounted. Collagen deposition was quantified using ImageJ software (National Institutes of Health). For each mouse, one trichrome-stained section from the injection site was analyzed. Within the implant region, three non-overlapping fields of view were selected, and a color-thresholding method was applied to segment aniline blue-positive collagen. The collagen-positive area was measured as a percentage of the total area for each field and averaged per section to yield an estimate of collagen density.
Immunofluorescence staining. Tissue sections were deparaffinized and rehydrated, followed by antigen retrieval using microwave heating in 1× Antigen Retrieval Buffer (prepared in 10% fetal bovine serum/PBS) for four cycles of 5 min each. After cooling to room temperature, slides were washed three times with 1× PBS (pH 7.4) for 3 min per wash and then blocked for 1 h at room temperature in 4% bovine serum albumin in PBS to minimize non-specific binding. Sections were incubated overnight at 4°C with primary antibodies against cluster of differentiation 86 (CD86; cat. #56-0862-82; 1:500; Invitrogen, Carlsbad, CA, USA) and cluster of differentiation 206 (CD206; cat. #36508; 1:2,000; Cell Signaling Technology, Danvers, MA, USA). Slides were subsequently rinsed with PBS, counterstained with 4′,6-diamidino-2-phenylindole, and mounted using an antifade mounting medium. Fluorescence images were acquired on a Zeiss LSM 710 or LSM 800 confocal microscope using ZEN software (Zeiss, Oberkochen, Germany). Fluorescence intensities of collagen I and collagen III were quantified using ImageJ (National Institutes of Health).
Statistical analysis. Data are reported as the mean ± standard error. GraphPad Prism 9 (GraphPad Software, Boston, MA, USA) was used for graphing and preliminary analyses, and statistical testing was conducted in SPSS v20 (IBM Corp., Armonk, NY, USA). Normality was evaluated using the Shapiro-Wilk test together with skewness and kurtosis. For PRIMOS volumetric measurements and histological outcomes, group differences at each time point were assessed using two-way analysis of variance followed by Bonferroni-adjusted post hoc tests for pairwise comparisons. Missing observations arising from scheduled euthanasia were accommodated using a mixed-model approach, which incorporates all available repeated measures without imputing values at later time points. A corrected value of p<0.05 was considered statistically significant.
Results
Characterization of filler particles. SEM was used to examine the morphology of PLLA and PDLLA fillers in powder form, after dispersion in saline, and after removal of the water-soluble excipient to isolate the PLA particles. SEM of the filler powders (Figure 1A) showed that both formulations contained spherical PLA cores partially embedded in a sheet-like excipient matrix. After dispersion in saline, PLLA microspheres retained smooth, well-defined shapes (Figure 1B upper panel), whereas PDLLA particles appeared distorted and fused into an irregular mesh (Figure 1B lower panel). Upon complete excipient removal, PLLA yielded intact, smooth spheres (Figure 1C upper panel), but PDLLA particles collapsed and coalesced into dense clusters (Figure 1C lower panel). These finding results indicate that PLLA fillers are larger and remain structurally stable through saline dispersion and excipient removal, preserving smooth, intact spherical particles.
Scanning electron microscopy images of poly-L-lactic acid (PLLA) and poly-D,L-lactic acid (PDLLA) fillers. (A) Filler powder, (B) filler dispersed in saline, and (C) poly-lactic acid particles after removal of the water-soluble excipient. Both filler types exhibited spherical poly-lactic acid cores associated with the excipient phase. PLLA maintained smooth, intact spheres after dispersion and excipient removal, whereas PDLLA particles were more easily deformed and collapsed into aggregated clusters after excipient removal.
Physicochemical and mechanical characterization of PLLA and PDLLA particles. Particle size analysis by SEM showed clear differences between the two materials. The PLLA particles exhibited a median diameter of 49.10 μm, with an approximately symmetric distribution (Figure 2A upper panel), whereas PDLLA particles were smaller, with a median diameter of 19.85 μm and a more right-skewed size distribution (Figure 2A lower panel).
Physicochemical and mechanical characterization of particles of poly-L-lactic acid (PLLA) (upper panel) and poly-D,L-lactic acid (PDLLA) (lower panel). (A) Size distributions of PLLA and PDLLA particles measured from scanning electron microscopy images. (B) Representative nuclear magnetic resonance spectra of particles isolated from PLLA and PDLLA fillers, showing spectral patterns consistent with the stereoregular (isotactic) structure of PLLA and the non-stereoregular (atactic) structure of PDLLA. (C) Differential scanning calorimetry thermograms of isolated PLLA and PDLLA particles. (D) Micro-compression test stress-strain curves of PLLA and PDLLA fillers measured in the presence of soluble excipients. (E) Micro-compression test stress-strain curves of isolated particles after deionized water washing/centrifugation to remove soluble excipients, showing different deformation and densification behaviors of PLLA and PDLLA.
The stereochemical composition of the isolated particles was confirmed by NMR. Spectra from the PLLA-derived particles were consistent with PLA composed solely of L-LA units, while particles isolated from the PDLLA filler showed signals attributable to PLA containing both L- and D-isomer units, consistent with a racemic lactide composition (Figure 2B). Thermal analysis further distinguished the materials (Figure 2C). PLLA particles displayed a glass transition (Tg=60.88°C) followed by a pronounced cold-crystallization event (Tcc=94.20°C; ΔHcc=−21.88 J/g) and clear melting endotherm (Tm=169.18°C; ΔHm=65.73 J/g), indicating higher crystallinity. In contrast, PDLLA particles exhibited no crystalline signals over the entire tested range and showed a lower glass transition temperature (Tg=55.04°C), consistent with a predominantly amorphous structure.
Micro-compression testing highlighted distinct deformation behaviors. In the presence of soluble excipients, PDLLA fillers deformed at lower applied pressures than PLLA across the strain range, indicating greater ductility (Figure 2D). After deionized water washing and centrifugation to remove soluble excipients, PDLLA particles still required lower pressure in the low-strain region (≤0.3). However, the pressure increased sharply at higher strains (Figure 2E), consistent with progressive densification/compaction of PDLLA during excipient removal. Overall, PLLA particles maintained a stiffer response with higher pressures required to reach comparable deformation.
These results indicate that PLLA particles exhibit greater thermal and mechanical robustness than PDLLA particles due to differences in their stereochemical composition.
Dispersion–aggregation–redispersion behavior in saline. When suspended at identical PLA concentration (42.5 mg/ml) in saline, PLLA and PDLLA fillers showed starkly different suspension stability (Figure 3A). PLLA suspensions remained relatively uniform, with only small clumps after agitation. In contrast, PDLLA rapidly formed visible flocs that persisted even after vortexing (Figure 3A). Quantitative circularity-based analysis (Figure 3B) confirmed that objects classified as fully aggregated clusters (circularity <0.1) had a median diameter of 380.0 μm for PLLA versus 458.1 μm for PDLLA. When normalized by their respective single-particle medians, of 49.10 μm for PLLA and 19.85 μm for PDLLA, these correspond to 7.7× and 23.1× increases in size, respectively (Figure 3B).
Dispersion–aggregation–redispersion behavior of poly-L-lactic acid (PLLA) and poly-D,L-lactic acid (PDLLA) fillers in phosphate-buffered saline as quantified by optical microscopy and circularity (Circ.) analysis. (A) Representative optical microscopy images of PLLA and PDLLA filler suspensions over a 3-day standing/agitation cycle. Scale bar=100 nm. (B) Size distribution of fully aggregated clusters identified by ImageJ circularity analysis (Circ. <0.1) for PLLA and PDLLA fillers. (C-E) Time-course of the degree of dispersion (%). The figure shows representative results obtained from a single experiment.
Consistent with these observations, the time-course of the degree of dispersion, defined by the fraction of objects exceeding circularity thresholds, remained higher for PLLA across all criteria (Figure 3C). For moderately stringent classification (circularity ≥0.4), PLLA maintained a high and stable dispersion level throughout the 3-day cycle, while PDLLA increased transiently after agitation but remained lower overall (Figure 3C). As the criterion became stricter (circularity ≥0.6 and ≥0.8), the disparity widened that PLLA retained a measurable population of near-spherical objects, whereas PDLLA showed consistently low values, indicating persistence of non-spherical, consolidated clusters even after vortexing (Figure 3D and E).
Together, these results indicate that PDLLA forms large, mechanically stable aggregates with poor re-dispersion, while PLLA is more readily re-dispersed into discrete entities under the same suspension and agitation conditions.
Volumizing effect and collagen deposition. After subcutaneous injection, both fillers exhibited an initial loss of apparent volume by week 4, corresponding to clearance of the HA. PRIMOS imaging (Figure 4A) showed that after the initial decline through week 4, projection volume exhibited a trend for partial rebound in both PLLA- and PDLLA-treated groups, whereas no such recovery pattern was observed in the PBS-treated group. The PLLA-treated group showed a gradual and relatively steady recovery trend on the volume-retention curve from week 4 to week 12, whereas the PDLLA-treated group showed a more delayed rebound pattern (Figure 4B). However, the within-group temporal change in PLLA-treated animals was not statistically significant, and therefore this pattern should be interpreted as a trend rather than a significant increase. Between weeks 4 and 12, the mean change in projection volume was 5.1 mm3 for the PLLA group and 4.8 mm3 for the PDLLA group.
In vivo volumetric retention and collagen deposition after subcutaneous injection of poly-L-lactic acid (PLLA) and poly-D,L-lactic acid (PDLLA) fillers. (A) Representative PRIMOS three-dimensional skin surface images of injection sites in nude mice at 0, 2, 4, 8, and 12 weeks after subcutaneous injection of phosphate-buffered saline (PBS), PLLA filler and PDLLA filler. (B) Quantification of projection volume over time as measured by PRIMOS for PBS, PLLA filler, and PDLLA filler (mean ± standard error). (C) Representative Masson’s trichrome-stained tissue sections harvested at the indicated time points showing tissue remodeling and collagen deposition around the injected materials. Collagen stained blue. Original magnifications, ×400. (D) Quantification of collagen-positive area (% area) from Masson’s trichrome-stained images at each time point (mean ± standard error). Significantly different at: *p<0.05, **p<0.01 and ***p<0.001 compared with the PBS control group using two-way analysis of variance.
Masson’s trichrome staining revealed more extensive collagen deposition in PLLA implants (Figure 4C). Over time, the PLLA-treated group showed a diffuse, homogeneous collagen network filling the implant space. Quantitatively, the PLLA-treated group had an approximately 3-fold higher collagen-positive area than the PDLLA-treated group at 2-8 weeks and 4-fold higher at 12 weeks (Figure 4D). In contrast, PDLLA-treated regions showed collagen primarily at the periphery of dense particle clusters, with much less staining internally (Figure 4C and D). The PBS control group had minimal collagen deposition except for a transient increase at week 4. Overall, these results suggest that the better dispersibility of PLLA allows fibroblasts to infiltrate broadly and deposit more collagen, whereas the aggregated morphology of PDLLA limits tissue access.
Histological analysis. Inflammatory cell infiltration and neovascularization at the injection sites were assessed on H&E-stained sections over 12 weeks by quantifying total inflammatory cells and the red blood cells area fraction. In the PBS-treated group, inflammatory cell counts remained low at all time points. In contrast, both PLLA and PDLLA fillers induced accumulation of inflammatory cells around the deposits (Figure 5A). The PDLLA group showed a pronounced inflammatory peak at 2 weeks, followed by a gradual decline thereafter. Across all time points, inflammatory cell counts were consistently lower in the PLLA-treated group than in the PDLLA-treated group (Figure 5B).
Inflammatory cell infiltration and neovascularization following subcutaneous injection of poly-L-lactic acid (PLLA) and poly-D,L-lactic acid (PDLLA) fillers. (A) Representative hematoxylin and eosin-stained sections of injection sites from phosphate-buffered saline (PBS), PDLLA, and PLLA groups at 0, 2, 4, 8, and 12 weeks, showing local tissue response and inflammatory cell infiltration. (B) Quantification of total inflammatory cells per high-power field over time. (C) Representative images of hematoxylin and eosin-stained sections highlighting neovascular structures at the injection sites. (D) Quantification of neovascularization expressed as red blood cells area fraction (% of tissue area). Data are presented as the mean ± standard error. Significantly different at: p<0.05, *p<0.01, and **p<0.001 by two-way analysis of variance with Bonferroni post hoc testing compared with PBS-treated group at the same time point.
Neovascularization was limited in all groups. The PBS-treated group showed no meaningful temporal change. The PDLLA-treated group exhibited a transient increase in neovascularization at 2 weeks, reaching >1.13% of the tissue area, and then decreased over time. By comparison, the PLLA-treated group showed consistently lower neovascularization than the PDLLA-treated group at all time points, with only a mild increase at 4 weeks, suggesting a more controlled vascular response at the injection site (Figure 5C and D).
These results showed an identical trend with the inflammatory cell infiltration results, suggesting that in the case of PDLLA, a greater number of neovessels were formed to support a more active foreign-body immune response.
Expression of M1 and M2 macrophages. Immunofluorescence staining was performed to assess macrophage polarization at the injection sites by measuring CD86 (M1) and CD206 (M2) signals across all time points. Representative images and quantitative analyses (Figure 6) showed that the fluorescence intensities of both CD86 and CD206 were significantly higher in the PDLLA-treated group, reaching approximately 3-fold greater levels than those observed in the PLLA-treated group. Collectively, these results indicate that PDLLA deposits induce a more pronounced and persistent inflammatory response than PLLA, as reflected by stronger macrophage-associated signals at the injection sites.
Macrophage polarization at injection sites following subcutaneous implantation of poly-L-lactic acid (PLLA) and poly-D,L-lactic acid (PDLLA) fillers. (A) Representative immunofluorescence images of showing macrophage-associated signals over time. Images were taken at ×200 magnification. Quantification of fluorescence intensities of CD86 (M1) (B) and CD206 (M2) (C) macrophages. 4′,6-Diamidino-2-phenylindole fluorescence is shown in blue, CD86 immunofluorescence is shown in green and CD206 immunofluorescence is shown in red. Data are presented as the mean ± standard error. Significantly different at: *p<0.05, **p<0.01, ***p<0.001, and ****p<0.0001 compared with the phosphate-buffered saline (PBS) control group using two-way analysis of variance with Bonferroni post hoc testing.
Discussion
This study shows that the stereochemistry of PLA profoundly influences filler behavior from the particle level to the tissue outcome. The PLLA formulation produced semicrystalline, mechanically rigid microspheres that remained individually discrete during handling and exposure to saline, whereas the PDLLA filler, an amorphous polymer, created deformable spheres that collapsed and coalesced into large aggregates. These material-level differences had clear in vivo consequences. Both fillers lost volume as HA and the water-soluble excipient fraction dissolved from week 0 to week 4, then regained volume as new matrix formed. However, the volume in the PLLA-treated group expanded steadily, while that of the PDLLA-treated group rebounded more slowly from week 4 to week 12. Crucially, PLLA stimulated broader collagen deposition with less inflammation, whereas PDLLA deposits remained highly inflamed but had relatively sparse collagen (Figure 4 and Figure 5).
Our NMR and DSC results confirmed the key material distinction that PLLA is stereoregular and can crystallize, whereas PDLLA has a random D,L-lactide composition, resulting in an amorphous structure. This compositional difference explains the mechanical data. As Lee et al. note, the amorphous polymer chains of PDLLA prevents rigid crystalline domains, making it less brittle and more ductile than PLLA (12). Correspondingly, in micro-compression, PDLLA particles deformed under low pressures, indicating high compliance. Conversely, the semicrystalline nature of PLLA rendered it stiffer and more resistant to deformation (22). Once we removed HA and water-soluble excipients, PDLLA underwent a distinct two-stage response, including easily compressible at first, and then stiffening as the collapsed structure densified (Figure 2E). In contrast, PLLA maintained consistently higher resistance across different strain levels. Thus, the inherently higher rigidity of PLLA spheres, due to their crystallinity, likely enabled them to survive agitation and shear without fracturing or compacting, whereas PDLLA spheres readily buckled and fused.
These physicochemical differences manifested at the macroscale in suspension. Under identical saline and agitation conditions, PDLLA suspensions rapidly aggregated meaning that the median PDLLA cluster was huge, 458 μm, relative to its single-particle size (median 19.85 μm), a 23-fold increase (Figure 3B). PLLA clusters were smaller, 380 μm (7.7-fold single-particle size), and re-dispersed more easily after vortexing (Figure 3C-E). Notably, PLLA retained a measurable population of near-spherical particles throughout the 3-day test, whereas PDLLA remained dominated by irregular clumps. Clinically, this matters because filler performance depends on how uniformly particles are distributed at the injection site and how much surface area they present to cells (23). Large aggregates can effectively reduce the accessible surface area and impose mass-transport constraints, which are known to modulate foreign-body responses and tissue integration around polymeric implants (24). Our histological findings support this. PDLLA deposits tended to appear as relatively dense agglomerates with less apparent, blue-stained collagen infiltration in Masson’s trichrome staining (Figure 4C), whereas PLLA deposits were associated with finer and more spatially distributed, blue-stained collagen deposition. In effect, better dispersibility acts as a biological ‘gatekeeper’ that yields a more permissive scaffold for cell infiltration and collagen neogenesis, while aggregation confines remodeling to the periphery (25).
The in vivo volumetric data align with these observations. Both fillers dropped precipitously in apparent volume by week 4, reflecting loss of the HA carrier. The similar early drop indicates that initial projection was largely due to the hydrated excipient, not the PLA spheres. The subsequent partial volume rebound is attributable to tissue remodeling around remaining PLA (26). The gradual volume increase by PLLA suggests that its dispersed microspheres allowed fibroblasts to infiltrate soon after week 4, depositing collagen uniformly (27). The delayed rebound by PDLLA implies that its dense clusters postponed interior remodeling until later, perhaps after partial breakdown or structural reorganization. Nevertheless, net gains by week 12 were comparable, indicating that both materials eventually achieved significant tissue augmentation, albeit with different kinetics and patterns.
Remarkably, PDLLA implants sustained higher inflammatory and macrophage signals despite lower collagen content. This inflammation-collagen decoupling suggests that the aggregated form of PDLLA elicits a strong immune response at exposed surfaces but restricts collagen synthesis where particles are consolidated. In effect, PDLLA may continuously attract macrophage cells, M1 or M2, to its outer shell (28). Consistent with this framework, dense agglomerates are expected to limit cell infiltration into the cluster core, confining remodeling to peri-aggregate regions (29). By contrast, milder inflammation induced by PLLA is spread across many particles, allowing collagen to form a broad network even though overall immune activation is lower (30). These data underscore that filler architecture, the geometry of the microsphere ensemble, is as important as polymer chemistry (31). A given material may provoke a certain magnitude of response, but its aggregation state determines how that response translates into tissue ingrowth.
Collectively, our findings have practical implications for filler design and handling. They suggest that achieving optimal outcomes with PLA fillers depends not only on the choice of PLLA or PDLLA, but on preserving particle integrity throughout preparation and injection. Formulation strategies, such as excipient choice, particle size distribution, and reconstitution protocols, such as mixing energy, dilution and filtration, should aim to minimize unwanted aggregation (8, 19). In vitro assays like that we used, tracking cluster formation and circularity in suspension, may serve as useful screening tools to predict in vivo behavior. Manufacturers and clinicians should be aware that an amorphous polymer like PDLLA may require more careful handling to prevent densification and nodularity. Conversely, PLLA formulations may be inherently more forgiving due to their crystalline toughness.
Study limitations. Our in vitro dispersion assay, while quantitative, used static 2D imaging and may not have captured all aspects of real injection such as needle shear, pressure during extrusion and 3D suspension rheology. The animal model was limited to 12 weeks. Longer-term studies are needed to link these early aggregation effects to ultimate filler longevity, degradation kinetics, and nodule formation risk. Finally, our comparison involved two proprietary products that differ in more than polymer stereochemistry such as HA formulation and particle porosity. Future work using synthetically controlled formulations, same HA carrier and matched particle sizes, would help isolate the independent impact of crystallinity and mechanics.
Conclusion
The data support a model in which the higher crystallinity and mechanical robustness of PLLA preserves microsphere integrity, sustains better dispersion and redispersion, and generates a more permissive microenvironment for fibroblast infiltration and homogeneous collagen neogenesis. The greater deformability of PDLLA promotes its collapse and consolidation into dense aggregates, which amplifies macrophage-dominant inflammation while restricting tissue ingrowth and collagen deposition within the implant region. These findings underscore dispersibility and particle integrity as central design parameters governing both volumizing performance and tissue remodeling outcomes in PLA-based dermal fillers.
Footnotes
Authors’ Contributions
Conceptualization: KMC, PNC, CYH. Data collection and curation: JYJ, CJJ, ZYD, SMK, YJP. Funding acquisition: CYH. Investigation: SMK, HSL. Methodology: JYJ, SMK, HSL, PNC. Project administration: KMC, PNC, CYH. Software: YXJ. Supervision: KMC, PNC, CYH. Validation: PNC, CYH. Writing–original draft: JYJ, CJJ, ZYD. Writing–review and editing: KMC, PNC, CYH.
Conflicts of Interest
SMK and HSL are employees of LabInCube Co. Ltd (Cheongju, Republic of Korea) which is developing the product. All other Authors declare no conflicts of interest.
Funding
This work was supported by the National Research Foundation of Korea(NRF) grant funded by the Korea government(MSIT) (No. RS-2022-NR070845). This research was supported by Industrial Technology Alchemist Project (No. 20025773, Development of Metropolitan Direct Air Capture and Utilization Technologies) funded by the Ministry of Trade, Industry & Energy (MOTIE, Korea).
Artificial Intelligence (AI) Disclosure
No artificial intelligence (AI) tools, including large language models or machine learning software, were used in the preparation, analysis, or presentation of this manuscript.
- Received January 29, 2026.
- Revision received March 15, 2026.
- Accepted March 19, 2026.
- Copyright © 2026 The Author(s). Published by the International Institute of Anticancer Research.
This is an open access article under the terms of the Creative Commons Attribution License, which permits use, distribution and reproduction in any medium, provided the original work is properly cited.
References
Jump to section
Related Articles
Cited By...
- No citing articles found.