Porosity-engineered long-acting protopanaxadiol microspheres prevent CDK4/6 inhibitor-induced myelosuppression Authors: Anan Zhang, Yunjing Zhu, Xinyu Kong, Zixu Liu, Mengxia Tan, Hongyun Yu, Guanghui Jing, Jingxin Gou, Tian Yin, Haibing He, Yu Zhang, Xuemei Zhang, Ke Liu, Peifu Xiao, Xing Tang Published: 2026 ABSTRACT AbstractCDK4/6 inhibitor-induced myelosuppression frequently necessitates treatment interruption or dose reduction, yet effective pharmacological interventions specifically addressing this persistent hematological toxicity remain limited. Here, we report a porosity-programmed protopanaxadiol-loaded PLGA microsphere (PPD-MS) system for rapid-onset and long-acting myeloprotection during anticancer treatment. Using a continuous-flow O/W microsphere fabrication strategy, microsphere porosity and core–shell architecture were precisely regulated by tuning aqueous-phase osmotic pressure, establishing a controllable process–structure–release relationship. The optimized PPD-MS rapidly achieved a stable plasma plateau within 0.5 h after intramuscular injection and maintained sustained systemic exposure for 19 days, with a strong in vitro–in vivo correlation between release profiles (R2 = 0.9983). Their narrow particle-size distribution enabled smooth administration through a 26G needle, supporting clinical feasibility. In chemotherapy- and CDK4/6 inhibitor-induced myelosuppression models, PPD-MS effectively restored peripheral blood cell counts, bone marrow cellularity, and hematopoietic stem and multipotent progenitor populations. Mechanistically, PPD-MS rescued hematopoietic cell-cycle progression by activating PI3K/AKT signaling, suppressing p21-mediated inhibition, and restoring the Rb/E2F1–Cyclin E/CDK2 axis, thereby reversing CDK4/6 inhibitor-induced G1/S arrest. Moreover, combination treatment with paclitaxel or CDK4/6 inhibition enhanced antitumor efficacy, accompanied by increased cytotoxic immune activation and reduced PD-L1-associated immune escape. Collectively, this study establishes a porosity-engineered long-acting microsphere platform that integrates predictable drug release, rapid and sustained myeloprotection, and improved combination anticancer therapy, providing a translational strategy for managing CDK4/6 inhibitor-associated hematological toxicity.Keywords: Protopanaxadiol-loaded PLGA microsphere, Porosity engineering, Long-acting drug delivery, Myelosuppression, CDK4/6 inhibitors, PI3K/AKT–Rb/E2F1 signaling Introduction Myelosuppression remains a common and clinically significant complication of cancer therapy. It can be induced by chemotherapy, radiotherapy, and an increasing number of small-molecule targeted agents [1,2]. In particular, cyclin-dependent kinase 4/6 inhibitors (CDK4/6i) are frequently associated with severe (grade ≥3) myelosuppression, with incidence rates reaching up to 90% in clinical studies; however, no established clinical interventions are currently available to effectively prevent or reverse this toxicity [3,4]. Once severe myelosuppression occurs, clinical management is largely limited to treatment interruption or dose reduction while awaiting spontaneous hematologic recovery [5]. This not only delays anticancer therapy but also increases the risk of life-threatening complications, including severe infections and hemorrhagic events, ultimately compromising patient prognosis [6]. Granulocyte colony-stimulating factor (G-CSF) is commonly used to manage chemotherapy- or radiotherapy-induced myelosuppression. G-CSF primarily acts on granulocytic progenitor cells, promoting their transition from the G0 phase to the G1 phase, thereby accelerating granulopoiesis and reducing the risk of infection [7]. However, CDK4/6i exert their pharmacological effects by blocking the transition from the G1 phase to the S phase of the cell cycle, which may partly explain the limited therapeutic benefit of G-CSF in CDK4/6i-induced myelosuppression [8]. Moreover, because G-CSF mainly targets granulocytic progenitor cells, long-term or repeated use may raise concerns regarding hematopoietic stem cell exhaustion and is often associated with adverse effects such as bone pain and fever. In addition, potential interference with anticancer efficacy has also been reported or suggested in certain therapeutic contexts [9]. Therefore, there is an urgent need to develop novel therapeutic strategies capable of effectively alleviating CDK4/6i-induced myelosuppression while maintaining favorable safety profiles and preserving, or even enhancing, antitumor efficacy. Protopanaxadiol (PPD), a major aglycone metabolite derived from Panax ginseng, possesses dual hematopoietic-regulatory and antitumor activities, making it particularly relevant to this therapeutic context [10]. Previous studies have demonstrated that PPD promotes hematopoietic recovery after bone marrow injury and exerts broad antitumor effects [11]. However, whether PPD can prevent CDK4/6i-induced myelosuppression and enhance the antitumor efficacy of CDK4/6i or chemotherapeutic agents remains largely unexplored. Moreover, the clinical translation of PPD has been severely hampered by its unfavorable pharmacokinetic properties, including poor aqueous solubility, rapid systemic clearance, low oral bioavailability, and a short in vivo half-life [12]. These limitations make it difficult to achieve stable and therapeutically effective plasma concentrations using conventional formulations. In clinical practice, agents used to manage myelosuppression are typically administered in synchrony with anticancer treatment cycles. Therefore, the development of a long-acting PPD formulation capable of single-dose administration per treatment cycle would significantly improve patient compliance and may contribute to shortening the intervals between chemotherapy cycles [13,14]. Long-acting injectable formulations based on biodegradable polymers, particularly poly(lactic-co-glycolic acid) (PLGA), offer a promising solution due to their favorable biocompatibility, biodegradability, and regulatory acceptance [15]. Furthermore, the porosity of PLGA microspheres can be precisely tuned to modulate the initial burst release and subsequent drug release kinetics [16,17]. Microfluidic fabrication techniques offer an attractive platform for achieving this goal, as they enable the production of highly uniform particles under well-defined and reproducible physicochemical conditions, thereby minimizing batch-to-batch variability. In addition, the continuous-flow nature of microfluidic processes facilitates scalable manufacturing with minimal scale-up effects during production [18]. In this study, we developed a microfluidics-enabled, porosity-engineered oil-in-water (O/W) microsphere fabrication platform, and successfully prepared PPD-loaded PLGA microsphere (PPD-MS) with precisely tunable internal microstructure and controlled drug release characteristics. By modulating the osmotic pressure of the aqueous phase during microfluidic fabrication, we achieved controlled regulation of microsphere porosity and early-stage drug release. The resulting PPD-MS provided rapid onset and sustained systemic exposure over an entire chemotherapy cycle, supported by a robust in vitro–in vivo correlation model. PPD-MS effectively alleviated both chemotherapy-induced acute myelosuppression and CDK4/6i-induced persistent hematopoietic suppression. Mechanistically, PPD-MS promoted hematopoietic stem cells (HSCs) and multipotent progenitors (MPPs) survival and restored cell-cycle progression by activating the PI3K/AKT-mediated p21–Rb/E2F1–Cyclin E/CDK2 axis. Furthermore, PPD-MS, when combined with chemotherapeutic agents or CDK4/6i, enhanced antitumor efficacy by activating the STAT1/NF-κB–Granzyme B/Perforin cytotoxic axis and suppressing the STAT3/PD-L1 immune escape axis (Scheme 1). Collectively, these findings establish PPD-MS as a process-controlled long-acting microsphere platform that enables predictable drug release and clinically translatable myeloprotection for CDK4/6i-associated hematological toxicity. Importantly, PPD-MS-based combination therapy not only mitigates treatment-related myelosuppression but also improves antitumor efficacy, highlighting its potential to improve the therapeutic index of anticancer treatment. Schematic illustration of PPD-MS and its dual functions in mitigating CDK4/6 inhibitor–induced myelosuppression and enhancing antitumor immunity. Materials and methods Materials PPD was supplied by Shandong Boyuan Biomedical Co., Ltd. (Yantai, China). PLGA (DLG 5050 2A) was provided by Shandong Luye Pharma Co., Ltd. (Yantai, China). Poly (vinyl alcohol) (PVA, 4–88), as well as HPLC-grade methanol and acetonitrile, were purchased from Merck KGaA (Darmstadt, Germany). Mannitol was obtained from Roquette (Lestrem, France). Ethyl acetate was purchased from Sinopharm Chemical Reagent Co., Ltd. (Nanjing, China). Propidium iodide (PI), ribonuclease A, YO-PRO-1 iodide, red blood cell lysis buffer, Dulbecco's Modified Eagle Medium (DMEM), and fetal bovine serum (FBS) were supplied by Dalian Meilun Biotechnology Co., Ltd. (Dalian, China). Recombinant anti-CD48 antibodies were obtained from Wuhan Sanying Biotechnology Co., Ltd. (Wuhan, China). Recombinant anti-CD150 antibodies were purchased from Wuhan Beinlai Biotechnology Co., Ltd. (Wuhan, China). Recombinant antibodies against CD4, CD8, Sca-1, CD117, CDK2, CDK4, CDK6, Cyclin D and Cyclin E were obtained from Wuhan Servicebio Technology Co., Ltd. (Wuhan, China). Sodium carboxymethyl cellulose was supplied by Ashland Specialties France (Biot, France). Cyclophosphamide (CTX) was purchased from Shanghai Yuanye Biotechnology Co., Ltd. (Shanghai, China). Palbociclib was obtained from Shanghai Energy Chemical Co., Ltd. (Shanghai, China). Paclitaxel (PTX) was supplied by Anhui Jinghe Biopharmaceutical Technology Co., Ltd. (Anhui, China). All other reagents were of analytical or chromatographic grade and used as received. Cells and animals The murine breast cancer cell line 4T1 was purchased from Dalian Meilun Biotechnology Co., Ltd. (Dalian, China). Cells were maintained in DMEM supplemented with 10% FBS and 1% penicillin–streptomycin at 37°C in a humidified atmosphere containing 5% CO2. Sprague–Dawley (SD) rats weighing 180–200 g, Institute of Cancer Research (ICR) mice weighing 18–20 g, and female BALB/c mice weighing 18–20 g were obtained from Jinan Pengyue Experimental Animal Breeding Co., Ltd. (Jinan, China). All animal experiments were conducted in strict accordance with the guidelines approved by the Animal Care and Use Committee of Shenyang Pharmaceutical University (Shenyang, China). Preparation of PPD-MS PPD-MS were prepared using a T-junction microfluidic method. The inner diameters of the oil and aqueous phase channels were 80 μm and 0.8 mm, respectively. Briefly, PPD and PLGA (40%:60%, w/w) were dissolved in ethyl acetate to form the oil phase, while 1% (w/v) PVA containing different concentrations of mannitol (0%, 4%, 10%, w/v) was used as the aqueous phase. The aqueous and oil phase flow rates were set at 110–310 μL s−1 and 2.5–6.5 μL min−1, respectively. The generated droplets were solidified by aqueous extraction under continuous stirring, collected through a 15 μm sieve, washed with purified water, and lyophilized to obtain PPD-MS. The final product was stored at 2-8°C. Particle size distribution and morphological characterization The morphology and size distribution of PPD-MS and emulsion droplets were observed using a digital microscope with ultra-high depth of field (Keyence, Japan) and a laser diffraction particle size analyzer (Mastersizer 3000, Malvern Panalytical Ltd., UK) [19]. Droplet shrinkage along the transport path was monitored to evaluate microsphere solidification [20]. The solidification efficiency (SE) was calculated as follows: where V0 is the initial droplet volume, Vt is the droplet volume at a given position along the tubing (corresponding to a specific solidification time), and Vlast is the final volume of the fully solidified PPD-MS. The overall morphology and surface characteristics of PPD-MS were examined by scanning electron microscopy (SEM; Phenom, Korea) [21]. Determination of drug loading and encapsulation efficiency of PPD-MS The actual drug loading (DL) and encapsulation efficiency (EE) of PPD-MS were determined as previously reported [[22], [23], [24]]. DL and EE were calculated according to the following equations: where WPPD is the actual amount of PPD encapsulated in the microspheres, WMS is the total mass of the dried microspheres, and WPPD0 is the theoretical amount of PPD added during preparation. Drug distribution and solid-state characterization of microspheres The spatial distribution of PPD within PPD-MS was characterized using confocal Raman microscopy (WITec GmbH, Ulm, Germany) [25]. The solid-state form of PPD in PPD-MS was evaluated by differential scanning calorimetry (DSC; Mettler-Toledo, Greifensee, Switzerland) [18]. Visualization of aqueous-phase permeability Fluorescent sodium was incorporated into the aqueous phase to enable visualization of the internal structure of wet microspheres. Undried PPD-MS were examined by confocal laser scanning microscopy (CLSM) (STELLARIS 5, Leica, Germany) using z-stack imaging [26]. Injectability and suspension stability of PPD-MS Non-uniform PPD-MS prepared by mechanical shearing were used as a control [27]. The injection force of PPD-MS suspension was measured at a constant rate of 5 mL min−1 using a texture analyzer (Stable Micro Systems, UK) [28]. Suspensions stability was evaluated using a multiple light scattering analyzer (Turbiscan Lab, Formulaction, France) [29]. In vitro drug release study The in vitro release behavior of PPD from PPD-MS was evaluated as in previous reports [30]. Briefly, 5 mg of PPD-MS was dispersed in 40 mL of release medium consisting of 0.01 M PBS containing 0.5% (w/v) Tween-80, adjusted to pH 7.4, and maintained at 37°C. The cumulative drug release percentage of PPD-MS was calculated using the following equation: where Ct is the PPD concentration at time t; V0 is the total release volume; Ci is the PPD concentration in each previously collected sample; V is the sampling volume; M0 is the mass of microspheres used. PPD-MS samples were collected at predetermined time points to examine morphological changes during in vitro release [31]. In vivo pharmacokinetic study To evaluate the in vivo release behavior of PPD-MS, five SD rats were intramuscularly injected with PPD-MS suspension into the right hind limb at a dose of 54 mg kg−1. Blood samples of approximately 300 μL were collected by retro-orbital sampling at designated time points. Plasma samples were processed by liquid–liquid extraction and quantified by liquid chromatography–tandem mass spectrometry (Thermo, USA) [24]. The cumulative in vivo release percentage of PPD was calculated using the trapezoidal method and the Wagner–Nelson equation: where Ct is the plasma drug concentration at time t, and Ke is the terminal elimination rate constant. Pharmacokinetic parameters were calculated using WinNonlin 6.4 software (Pharsight, Certara Corporation, St. Louis, USA). IVIVC was evaluated by linear regression analysis between the cumulative in vivo and in vitro release percentages [28,32]. Therapeutic effect of PPD-MS on CTX-induced acute myelosuppression To evaluate the therapeutic effects of PPD-MS against chemotherapy-induced acute myelosuppression, 25 ICR mice were randomly divided into five groups. Mice in four groups were intraperitoneally administered CTX at a dose of 150 mg kg−1 in normal saline to establish an acute chemotherapy-induced myelosuppression model [33]. The healthy control group received an equal volume of normal saline via intraperitoneal injection. Mice in the low-, medium-, and high-dose PPD-MS groups were intramuscularly injected with PPD-MS at doses of 80, 120, and 160 mg kg−1, respectively, 1 day before CTX administration. Mice in model group and normal group received intramuscular injections of an equal volume of normal saline. At designated time points, blood cell counts were analyzed using a HemaVet 950FS Hematology Analyzer (Drew Scientific Inc., USA). On day 6 after CTX administration, mice were euthanized, and femurs and tibias from the hind limbs were harvested. A portion of the bone tissues was fixed in paraformaldehyde for histological and immunofluorescence analyses. H&E staining was performed for morphological evaluation, and immunofluorescence staining was conducted for CD117, Sca-1, CD48, and CD150 to identify hematopoietic stem and progenitor cell subsets [34]. Preventive effect of PPD-MS on CDK4/6i–induced persistent myelosuppression To evaluate the preventive effects of PPD-MS against CDK4/6i-induced persistent myelosuppression, 25 ICR mice were randomly divided into five groups [35]. Four groups were orally administered palbociclib suspension at 75 mg kg−1 once daily to establish a CDK4/6i–induced persistent myelosuppression model, while normal group received an equal volume of normal saline by oral gavage. PPD-MS was intramuscularly administered to the low-, medium-, and high-dose groups at 80, 120, and 160 mg kg−1, respectively, on the first day of palbociclib administration. Blood cell counts were analyzed using a HemaVet 950FS Hematology Analyzer. After 10 days of treatment, mice were euthanized, and femurs and tibias were harvested. A portion of the bone samples was fixed in paraformaldehyde for H&E staining and immunofluorescence analysis. The remaining samples were used for bone marrow cell isolation and enrichment, followed by flow cytometric analysis of bone marrow cell-cycle distribution and Western blot (WB) analysis of related proteins [31]. Antitumor efficacy of PPD-MS in the breast cancer model To evaluate the antitumor effects of PPD-MS alone and in combination with PTX or a CDK4/6i on breast cancer progression, 30 BALB/c mice were subcutaneously inoculated with 100 μL of 4T1 cell suspension at 1 × 107 cells·mL−1 to establish a syngeneic breast cancer model [36]. When tumor volume reached approximately 100 mm3, mice were randomly divided into six groups (n = 5 per group): model, PPD-MS, PTX, PPD-MS + PTX, CDK4/6i, and PPD-MS + CDK4/6i. The model control group received an intramuscular injection of normal saline on day 1. The PPD-MS group received a single intramuscular injection of PPD-MS at 120 mg kg−1 on day 1. The PTX group received intraperitoneal PTX at 10 mg kg−1 every 3 days. The CDK4/6i group received oral palbociclib suspension at 75 mg kg−1 once daily. Combination groups received the corresponding combined regimens. Tumor volume and body weight were measured every 2 days. At the end of treatment, mice were euthanized and tumor, thymus, and spleen were collected, weighed, and processed for cell cycle analysis, immunohistochemistry, transcriptomic profiling, and WB analysis. Tumor volume was calculated as follow [37]: where a and b represent the longest and shortest diameter, respectively. In vitro cytotoxicity assay Drug interaction was evaluated using the combination index (CI) according to the Chou-Talalay method [38]. Tumor cells were treated with PPD, PTX, CDK4/6i, or their combinations for 24 h, and cell viability was determined by CCK-8 assay. CI values were calculated based on the resulting dose-response data, where CI < 1, CI = 1, and CI > 1 indicate synergistic, additive, and antagonistic effects, respectively. Safety evaluation of PPD-MS 10 ICR mice were randomly divided into two groups. Mice in the PPD-MS group received a single intramuscular injection of PPD-MS at a dose of 320 mg kg−1, while mice in the saline control group were intramuscularly injected with an equal volume of normal saline. Body weight variations of the two groups of mice were monitored, followed by serum biochemical assessment and histopathological evaluation [39]. Statistical analysis Statistical analyses were performed using GraphPad Prism software (version 8.0.1). All data are presented as the mean ± standard deviation (mean ± S.D.). Comparisons among multiple groups were conducted using one-way analysis of variance (ANOVA). A value of P < 0.05 was considered statistically significant. Statistical significance in figures was denoted as follows: *P < 0.05, **P < 0.01, ***P < 0.001, and ****P < 0.0001. Results and discussion Microfluidic preparation and size-controlled fabrication of PPD-MS The preparation process of PPD-MS is illustrated in Fig. 1A. A quantitative relationship between the aqueous-phase flow rate and the particle sizes of both the primary emulsion droplets and the resulting PPD-MS was established (Fig. 1B–S1). A linear correlation model between the primary emulsion droplet size and the final PPD-MS particle size was successfully established (R2 = 0.9846, Fig. 1C), enabling real-time prediction of the final microsphere size during preparation. The resulting PPD-MS exhibited high particle size uniformity, with span values below 0.15 (Fig. 1D and E). By establishing a rapid microsphere solidification process, more than 80% droplet solidification was completed within 2 s at aqueous phase flow rates of 110–280 μL s−1, thereby ensuring efficient PPD incorporation into PPD-MS, as reflected by a high actual drug loading of 39.98% and an encapsulation efficiency of up to 99.94% (Fig. 1F–S2-S4) [40,41]. The established continuous microsphere fabrication platform exhibits broad applicability and is readily adaptable to the preparation of oil-in-water type microspheres, including risperidone sustained-release microspheres, protopanaxatriol sustained-release microspheres, and paliperidone palmitate sustained-release microspheres (Figs S5–S7). During 12 months of storage, PPD-MS remained a white, free-flowing powder without aggregation or caking. Its drug loading (38.82%), particle size, and in vitro release profile (f2 = 77%) remained essentially unchanged from their initial values (Figs. S8–S9). Preparation process and physicochemical characterization of PPD-MS. (A) Schematic illustration of preparation process of PPD-MS; (B) Correlation between aqueous phase flow rate and the droplet size as well as the final microsphere size of PPD-MS. (C) Linear correlation analysis between the emulsion droplet size and the size of PPD-MS (n = 3); (D) Particle size distribution profiles of non-uniform PPD-MS and size-uniform PPD-MS (n = 3); (E) SEM image of PPD-MS (scale bar = 80 μm); (F) Solidification efficiency under different aqueous-phase flow rates; (G) Raman mapping image of PPD-MS (scale bar = 10 μm, Red is PPD, Green is PLGA.); (H) DSC thermograms of PPD-MS and the corresponding raw materials. Confocal Raman mapping revealed that PPD was encapsulated by PLGA layers and localized in the interior of the microspheres, forming a typical concentric core–shell structure (Fig. 1G) [42]. DSC analysis further demonstrated that PPD predominantly existed in a crystalline state within the microspheres (Fig. 1H) [43]. The formation of this concentric core–shell structure was mainly attributed to the marked solubility difference between PPD and PLGA in the organic solvent system, as well as the rapid and highly efficient solidification process [44]. Compared with previously reported PLGA microsphere systems in which drugs are molecularly dispersed or preferentially enriched near the particle surface, the PLGA shell-encapsulated crystalline drug structure observed in this study is more effective in suppressing the initial burst release [45]. Consequently, this unique concentric core–shell architecture provides a structural basis for the long-acting and sustained release performance of PPD-MS [46]. Porosity engineering and underlying formation mechanism of PPD-MS To enable PPD-MS to rapidly attain therapeutically relevant plasma concentrations after administration, we sought to regulate the final porosity of PPD-MS by controlling water diffusion into emulsion droplets, thereby modulating the drug release rate (Fig. 2A) [47,48]. The water content of wet microspheres decreased with increasing aqueous phase osmotic pressure (Fig. 2B). Correspondingly, microspheres prepared at lower osmotic pressure exhibited larger particle sizes despite similar primary droplet sizes (Fig. 2C), suggesting that particle size was mainly influenced by internal porosity. Confocal microscopy revealed progressively reduced aqueous phase penetration as osmotic pressure increased, evidenced by a thinner fluorescent layer within the microspheres (Fig. 2D and E). This observation was further confirmed by cross-sectional morphology, which showed decreased porosity at higher osmotic pressures (Fig. 2F and G) [49]. Collectively, these results demonstrate that the aqueous-phase osmotic pressure is negatively correlated with the porosity of PPD-MS. By modulating the osmotic pressure of the aqueous phase, PPD-MS with tailored porosity can be readily prepared, providing a practical and controllable strategy for integrated formulation process–structure design. Characterization of pore structure, in vitro release behavior and injectability performance of PPD-MS (A) Schematic illustration of the pore formation mechanism of PPD-MS; (B) Water content of PPD-MS wet microspheres (n = 3); (C) D50 particle size of PPD-MS microspheres; (D) CLSM images of fluorescein sodium-labeled PPD-MS wet microspheres; (E) Penetration depth of fluorescein sodium within wet microspheres; (F) SEM images of cross-sections of PPD-MS (scale bar = 20 μm); (G) Correlation between aqueous-phase osmotic pressure and the final porosity of PPD-MS; (H) Korsmeyer–Peppas release exponent (n) and cumulative drug release at 24 h for PPD-MS with different porosities; (I) SEM images of high-porosity PPD-MS during the drug-release process (scale bar = 20 μm); (J-L) Glide force of PPD-MS during injection through needles of 21G (J), 23G (K), 26G (L); (M) Suspension stability profiles of PPD-MS. The in vitro release kinetics demonstrated that PPD-MS porosity was closely associated with its release behavior (Fig. 2H–S10). In particular, the release exponent n derived from the Korsmeyer–Peppas (K–P) model, indicated that increased porosity facilitated early-stage PPD diffusion, thereby accelerating the initial release rate from the microspheres (Fig. 2H–S11–S13). To clarify the porosity-dependent release mechanism of PPD-MS, the morphology of high-porosity PPD-MS were monitored during release. As shown in Fig. 2I, pores rapidly developed on the microsphere surface at the early stage and subsequently decreased or partially closed. These transient pores shortened the diffusion pathway of PPD within the core–shell structure, enabling rapid drug diffusion and providing an initial loading dose [50]. These results indicate that aqueous-phase osmotic pressure can be used as an effective process parameter to regulate the initial burst release and overall release kinetics of PPD-MS, thereby providing a structural basis for rapid onset and sustained therapeutic efficacy after administration [51]. Furthermore, high-porosity PPD-MS exhibited improved suspension stability (Fig. 2M), and the highly uniform microspheres could be smoothly administered through a 26G needle (Fig. 2J–L). This finding is consistent with Stokes’ equation, as increased porosity may reduce the effective particle density and thereby decrease sedimentation velocity [52]. Clinically, the enhanced suspension stability of high-porosity PPD-MS provides a longer handling window before injection, while compatibility with a small-gauge needle may help reduce injection-associated pain and anxiety in patients [28]. In vivo pharmacokinetics and IVIVC To comprehensively assess the in vivo release behavior and systemic exposure profile of PPD-MS, its pharmacokinetic characteristics were first evaluated in SD rats. As shown in Fig. 3A, following intramuscular administration of high-porosity PPD-MS, plasma PPD concentrations rapidly reached a stable plateau within 0.5 h and were sustained for up to 19 days. Plasma exposure remained highly stable over the entire dosing interval, with a steady-state deviation of only 22.72 ng mL−1. These results indicate that high-porosity PPD-MS provides both rapid onset and long-lasting, stable systemic exposure, enabling single-injection coverage over an entire chemotherapy cycle. In vivo release characteristics and IVIVC of PPD-MS. (A) Plasma concentration–time profile of PPD-MS (n = 5); (B) Comparison of cumulative drug release profiles of PPD-MS in vitro and in vivo; (C) IVIVC curve of PPD-MS release behavior; (D–I) In vivo release kinetic model fitting of PPD-MS, including (D) zero-order kinetic model, (E) first-order kinetic model, (F) Higuchi model, (G) Korsmeyer–Peppas model, (H) Peppas–Sahlin model, (I) Weibull model; (J) Schematic diagram of the release mechanism. The cumulative in vivo fraction released was subsequently calculated using the Wagner–Nelson method (Fig. 3B) and compared with the corresponding in vitro release profile. As shown in Fig. 3C, an excellent correlation was obtained, with a coefficient of determination (R2) of 0.9983, demonstrating a strong IVIVC [28]. To further evaluate the predictive performance of the established IVIVC model, the in vivo cumulative release profiles were back-calculated using the Wagner–Nelson equation and compared with the model-predicted profiles. The mean residual of the in vivo cumulative release was 3.15%, with a maximum residual of 12.83%, which was below 15%. In addition, the prediction errors for AUC and Cmax were 3.21% and 0.66%, respectively. These results demonstrate that the established IVIVC model possessed satisfactory predictive accuracy for in vivo drug release and systemic exposure. To further elucidate the mechanistic basis underlying the favorable in vivo release behavior of PPD-MS, several classical kinetic models were applied to fit the in vivo release profile (Fig. 3D–I). Comparison based on the coefficient of determination revealed that the K–P model provided the best fit (R2 = 0.9976, Fig. 3G), with a release exponent n of 0.63, indicating a combined diffusion-and-erosion mechanism (0.43 < n < 0.85) [50]. Further analysis using the Peppas–Sahlin model identified a crossover time at t = 38.99 h, at which R/F = 1 (Fig. 3H). Consistent with the in vitro morphological evolution results, these findings support a time-segmented release mechanism, in which early-stage release is predominantly governed by porosity-driven diffusion, followed by a later-stage release phase dominated by matrix erosion (Fig. 3J) [53]. Therapeutic effects of PPD-MS on chemotherapy-induced acute myelosuppression To further evaluate the therapeutic effects of PPD-MS on chemotherapy-induced acute myelosuppression, peripheral blood cell counts were monitored at different time points after model establishment (Fig. 4A). As shown in Fig. 4B, after 4 days of treatment, mice receiving medium- and high-dose PPD-MS exhibited significantly increased WBC and NEUT counts compared with the model group (P < 0.05). By day 6, all PPD-MS treatment groups exhibited significant recovery in WBC, NEUT, and MONO counts compared with the model group (P < 0.05). Notably, in the high-dose group, the numbers of WBCs, NEUTs, MONOs, and LYMs were increased by 4.3-, 3.5-, 4.3-, and 6.3-fold, respectively, relative to those in the model group. These results demonstrate that PPD-MS effectively alleviates chemotherapy-induced leukopenia and exerts a dose-dependent therapeutic effect within the dose range of 80–160 mg/kg (P = 0.0009). Therapeutic effects of PPD-MS on CTX-induced acute myelosuppression. (A) Schematic diagram of the treatment process; (B) Changes in peripheral blood cell counts in mice from different treatment groups; (C) Cell cycle distribution of bone marrow cells from different groups analyzed by flow cytometry; (D) Percentage of bone marrow cells in S phase and G2/M phase; (E) mRNA expression levels of Cyclin A, Cyclin B1 in bone marrow cells; (F) Flow cytometric analysis of apoptosis in bone marrow cells from different treatment groups; (G) Percentage of bone marrow cells undergoing early and late apotosis; (H) mRNA expression levels of Bcl-XL and Bcl-2 in bone marrow cells; (I) H&E staining of bone marrow (scale bar = 50 μm); (J) Immunofluorescence images showing LT-HSCs, ST-HSCs, MPP2, and MPP3/4 in bone marrow (scale bar = 50 μm); (K) Semi-quantitative analysis of the densities of LT-HSCs, ST-HSCs, and MPP2/3/4 populations in the bone marrow. (n = 5, *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001). Flow cytometric analysis of bone marrow cell-cycle distribution (Fig. 4C and D) revealed that PPD-MS effectively alleviated the CTX-induced cell-cycle arrest of bone marrow cells at the S-to-G2/M phase transition, with the high-dose PPD-MS group restoring the proportions of S- and G2/M-phase cells to near-normal levels. Mechanistically, PPD-MS significantly upregulated the expression of Cyclin A and Cyclin B1 in a dose-dependent manner (Fig. 4E, P < 0.05), thereby mitigating CTX-induced cell-cycle inhibition [54]. Apoptosis analysis by flow cytometry demonstrated that PPD-MS markedly antagonized CTX-induced early- and late-stage apoptosis of bone marrow cells (Fig. 4F and G, P < 0.05), exhibiting a strong dose-dependent protective effect (P < 0.0001). Mechanistically, PPD-MS treatment significantly restored the expression of the anti-apoptotic proteins BCL-2 and BCL-XL (P < 0.05), thereby counteracting CTX-induced apoptosis (Fig. 4H) [54]. Histopathological examination by H&E staining revealed that PPD-MS treatment markedly increased the density of nucleated bone marrow cells compared with the CTX model group (Fig. 4I). Moreover, immunofluorescence analysis demonstrated that PPD-MS treatment significantly restored HSCs and MPPs populations, including long-term HSCs (LT-HSCs), short-term HSCs (ST-HSCs), MPP-2 and MPP-3/4 (Fig. 4J and K) [34]. In contrast to G-CSF, which has been associated with excessive mobilization of MPPs and potential hematopoietic stem cell exhaustion, PPD-MS exhibits superior potential clinical advantages against chemotherapy-induced acute myelosuppression [55]. Prevention of CDK4/6i-induced persistent myelosuppression by PPD-MS Given the pronounced protective effects of PPD-MS against chemotherapy-induced acute myelosuppression, the present study further investigated its preventive potential against CDK4/6i induced persistent myelosuppression (Fig. 5A). As shown in Fig. 5B, PPD-MS effectively prevented the reductions in WBC, NEUT, LYM, MONO, and PLT induced by continuous oral administration of CDK4/6i (P < 0.05). Notably, in the medium- and high-dose PPD-MS groups, the counts of WBCs, NEUTs, MONOs, and PLTs were comparable to those in the normal control group (P > 0.05). Moreover, relative to the model group, the high-dose PPD-MS group achieved recovery rates of 87.89%, 205.00%, 38.35%, 89.88%, and 105.26% for WBC, NEUT, LYM, MONO, and PLT, respectively. Preventive effects of PPD-MS on CDK4/6i–induced persistent myelosuppression. (A) Schematic diagram of the treatment process; (B) Changes in peripheral blood cell counts in mice after cumulative oral administration of CDK4/6i for 10 days under different treatment conditions; (C) Cell cycle distribution of bone marrow cells from different groups analyzed by flow cytometry; (D) Percentage of bone marrow cells in G1 phase and S phase; (E) H&E staining of bone marrow (scale bar = 50 μm). Flow cytometric analysis of bone marrow cells further revealed that PPD-MS effectively alleviated CDK4/6i-induced G1-to-S phase cell-cycle arrest (Fig. 5C and D). Consistently, H&E staining of bone marrow sections provided direct morphological evidence that co-administration of PPD-MS markedly restored bone marrow cellularity (Fig. 5E). In addition, immunofluorescence analysis demonstrated that PPD-MS significantly reversed the CDK4/6i–induced reduction in MPPs populations, including MPP-2 and MPP-3/4 (Fig. 6A and B). Together, these findings suggest that PPD-MS restores the CDK4/6i–induced blockade of proliferation and differentiation in ST-HSCs, MPP-2, and MPP3/4, thereby alleviating persistent myelosuppression [56]. Mechanistic study on the preventive effect of PPD-MS against CDK4/6i–induced persistent myelosuppression. (A) Immunofluorescence images of LT-HSCs, ST-HSCs, and MPP2, MPP3, and MPP4 in mouse bone marrow (scale bar = 50 μm); (B) Quantitative analysis of the densities of LT-HSCs, ST-HSCs, and MPP2/3/4 populations in the bone marrow; (C-G) Quantitative analysis of CDK2 (C), Cyclin E (D), CDK4 (E), CDK6 (F) and Cyclin D (G) expression in bone marrow; (H) Representative Western blot images; (I) Quantification of relative protein expression levels; (J) Schematic illustration of the protein mechanism. (n = 5, *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001). To further clarify the molecular mechanisms underlying the protective effects of PPD-MS against CDK4/6i-induced myelosuppression, the expression of G1-to-S phase transition-related proteins in bone marrow cells was analyzed (Fig. 6C–G, S14). Immunofluorescence analysis initially showed that PPD-MS activated the Cyclin E–CDK2 axis by upregulating CDK2 and Cyclin E, thereby enabling bone marrow cells to overcome the G1 restriction point and restore their proliferative and differentiative capacity [57]. Further Western blot analysis revealed that PPD-MS activated PI3K/AKT-related signaling, suppressed p21-mediated cell-cycle inhibition, restored Cyclin E/CDK2 activity, and promoted Rb phosphorylation and E2F1 activation, collectively rescuing the G1/S transition and preventing CDK4/6i-induced myelosuppression (Fig. 6H and I) [58]. Collectively, these findings demonstrate that PPD-MS prevents CDK4/6i-induced persistent myelosuppression by restoring G1/S cell-cycle progression through activation of the PI3K/AKT–p21–Rb/E2F1–Cyclin E/CDK2 axis, highlighting its strong translational potential as a promising strategy to address the unmet clinical need for effective interventions against CDK4/6i-associated hematological toxicity (Fig. 6J) [59]. Immunomodulatory antitumor efficacy of PPD-MS alone and in combination therapy Given the demonstrated efficacy of PPD-MS in antagonizing both acute and persistent myelosuppression, the antitumor effects of PPD-MS were next systematically evaluated, either as monotherapy or in combination with PTX or a CDK4/6i (Fig. 7A). As shown in Fig. 7B, treatment with PPD-MS, PTX, PPD-MS + PTX, CDK4/6i, or PPD-MS + CDK4/6i significantly suppressed tumor growth compared with the saline group (P < 0.05). Notably, PPD-MS in combination with either PTX or CDK4/6i exhibited significantly greater antitumor activity observed in the combination treatment groups (P < 0.05). Consistent with tumor volume changes, tumor weight ratios and tumor growth inhibition further confirmed the enhanced antitumor efficacy of the combination treatments (Fig. 7C and E, and S15) [36]. Gross tumor images, H&E staining, and TUNEL assays provided additional evidence that PPD-MS, either alone or in combination with PTX or CDK4/6i, markedly suppressed tumor growth and promoted tumor cell apoptosis (Fig. 7D, 7F-7H) [36]. Antitumor efficacy of PPD-MS alone or in combination with PTX and CDK4/6i in a 4T1 tumor-bearing mouse model. (A) Schematic diagram of the treatment process; (B) Tumor volume growth curves; (C) Body weight changes; (D) Representative images of excised tumors; (E) Tumor inhibition rates; (F) H&E staining of tumor tissue sections (scale bar = 50 μm); (G) TUNEL staining of tumor sections (scale bar = 50 μm); (H) Quantitative analysis of apoptotic rates in tumor tissues based on TUNEL staining. (n = 5, *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001). To elucidate the mechanistic basis underlying the antitumor activity of PPD-MS, transcriptomic analyses were performed on tumor tissues, and gene expression profiles were systematically compared between groups with and without PPD-MS treatment. As shown in Fig. 8A and S16, additional PPD-MS treatment induced extensive transcriptional alterations, including both upregulated and downregulated genes. Gene Ontology (GO) enrichment analysis, Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway analysis, and gene set enrichment analysis (GSEA) consistently revealed significant enrichment of multiple immune activation-related pathways following PPD-MS treatment (Fig. 8B and C, and S17) [60]. These findings suggest that PPD-MS exerts antitumor effects, either as monotherapy or in combination with PTX or CDK4/6i, at least in part by modulating host antitumor immune responses [61]. Transcriptomic analysis of tumor tissues following PPD-MS treatment. (A) Volcano plots of differentially expressed genes in tumor tissues; (B) KEGG pathway enrichment analysis; (C) GSEA pathway enrichment analysis; (D-E) H&E staining of spleen (D, scale bar = 100 μm) and thymus (E, scale bar = 50 μm). In addition, PPD-MS treatment significantly reduced the tumor-associated increase in thymus and spleen indices (P < 0.05). H&E staining of the thymus and spleen further confirmed that PPD-MS alleviated tumor-associated pathological alterations in these immune organs (Fig. 8D, E, 9A, and 9B). Immunomodulatory mechanisms underlying the antitumor effects of PPD-MS. (A-B) Thymus (A) and Spleen (B) indices of mice; (C-D) Immunofluorescence staining images of CD4+ (C) and CD8+ (D) T cells in tumor tissues (scale bar = 50 μm); (E-F) Quantitative analysis of CD4+ (E) and CD8+ (F) T cell–positive populations in tumor tissues; (G-H) Serum levels of IL-6 (G) and IFN-γ (H); (I) Representative Western blot images; (J) Quantification of relative protein expression levels; (K) Schematic illustration of the protein mechanism. (n = 5, *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001). Immunofluorescence analysis of tumor tissues revealed that PPD-MS markedly increased infiltration of CD4+ T cells and CD8+ T cells (Fig. 9C–F). Given that CDK4/6i have been reported to modulate the tumor immune microenvironment through effects on regulatory T cells (Tregs), the increased infiltration of effector T cells observed in the present study may reflect a shift toward a more immunostimulatory microenvironment [62]. However, Treg populations were not evaluated and warrant further investigation. Concurrently, serum cytokine analysis demonstrated that combined PPD-MS treatment markedly elevated IL-6 and IFN-γ levels relative to the saline, PTX or CDK4/6i groups (Fig. 9G and H) [63]. Further Western blot analysis showed that PPD-MS upregulated p-STAT1, NF-κB, Granzyme B, and Perforin, while downregulating p-STAT3 and PD-L1 (Fig. 9I and J). These molecular changes indicate that PPD-MS promotes antitumor immune activation, enhances cytotoxic T cell/NK cell-mediated killing, and suppresses PD-L1-mediated tumor immune escape [64]. Interestingly, although PPD-MS combined with PTX or CDK4/6i produced superior tumor suppression in vivo compared with either monotherapy, CI analysis at the cellular level demonstrated antagonistic interactions. The CI values of PPD + PTX and PPD + CDK4/6i at Fa = 0.5 were 1.28 and 1.50, respectively, indicating that the enhanced therapeutic efficacy observed in vivo was not attributable to direct pharmacological synergy (Figure S18). This observation is consistent with previous reports showing that CDK4/6i may antagonize the cytotoxic activity of chemotherapeutic agents by inducing G1 cell-cycle arrest, thereby reducing the susceptibility of tumor cells to cell cycle-dependent cytotoxic drugs [65]. Therefore, the improved antitumor efficacy observed in vivo is unlikely to arise from enhanced direct tumor cell killing. Instead, the superior therapeutic outcome may be associated with the immunomodulatory effects of PPD-MS. The transcriptomic analyses, increased CD4+ and CD8+ T-cell infiltration, elevated IFN-γ production, and regulation of STAT1/NF-κB and STAT3/PD-L1 signaling collectively suggest that PPD-MS promotes antitumor immune responses, which may contribute to the enhanced efficacy of combination treatment in vivo (Fig. 9K) [66]. Systemic safety and local biocompatibility of PPD-MS The safety assessment was conducted using PPD-MS at twice the maximum therapeutic dose. As shown in Fig. 10A, no significant changes in body weight were observed over 16 days after PPD-MS administration compared with healthy control mice (P > 0.05). Serum biochemical analyses further demonstrated that biomarkers of hepatic, cardiac, and renal function in the PPD-MS group were comparable to those in healthy controls (P > 0.05, Fig. 10B–L). Consistently, H&E histopathological examination revealed no obvious pathological alterations in the heart, liver, spleen, lungs, or kidneys after PPD-MS treatment (Fig. 10M) [39]. In vivo injection safety evaluation of PPD-MS. (A) Body weight changes of mice; (B ∼ L) Serum biochemical parameters related to liver, kidney, and cardiac function in mice; (M) H&E staining of heart, liver, spleen, lung, kidney, and injection-site muscle tissues (scale bar = 50 μm). (n = 5, ns, P > 0.05). Histological analysis of muscle tissue at the injection site showed residual microspheres accompanied by localized macrophage infiltration, while the surrounding muscle fibers remained intact, with no evidence of necrosis or degenerative changes (Fig. 10M). These findings are characteristic of PLGA-based microsphere formulations and reflect the favorable local biocompatibility of PLGA materials [39,67]. Collectively, these results demonstrate that PPD-MS exhibits favorable systemic safety and local tissue compatibility, even at twice the maximum therapeutic dose, supporting its suitability for long-acting intramuscular administration. Conclusion To address the critical unmet clinical need for effective interventions against CDK4/6i-induced myelosuppression, this study established a long-acting PPD-MS as an integrated therapeutic strategy. PPD-MS prevented CDK4/6i-induced persistent myelosuppression by restoring G1/S cell-cycle progression through the PI3K/AKT–p21–Rb/E2F1–Cyclin E/CDK2 axis and preserving MPP-2 and MPP-3/4 populations. Meanwhile, PPD-MS alleviated chemotherapy-induced acute myelosuppression through upregulation of Cyclin A, Cyclin B1, Bcl-2, and Bcl-xL, while preserving HSCs and MPPs populations. Furthermore, PPD-MS enhanced the antitumor efficacy of chemotherapeutic agents and CDK4/6i by activating the STAT1/NF-κB–Granzyme B/Perforin cytotoxic axis and suppressing the STAT3/PD-L1 immune escape axis. Patient compliance and clinical operability were also carefully considered in the formulation design of PPD-MS. The unique core–shell and high-porosity structure enabled rapid attainment of a stable plasma concentration plateau within 0.5 h and sustained drug release over an entire chemotherapy cycle, supporting once-per-cycle administration. Moreover, the uniform particle size and high porosity of PPD-MS ensured injectability through a small-gauge needle, potentially reducing injection-related pain and anxiety. Collectively, these findings establish PPD-MS as a porosity-engineered long-acting microsphere platform that provides rapid-onset and sustained myeloprotection while enhancing antitumor efficacy, offering a promising translational strategy for managing therapy-associated hematological toxicity. CRediT authorship contribution statement Anan Zhang: Writing – original draft, Visualization, Software, Methodology. Yunjing Zhu: Software. Xinyu Kong: Resources. Zixu Liu: Writing – review & editing. Mengxia Tan: Visualization. Hongyun Yu: Visualization. Guanghui Jing: Investigation. Jingxin Gou: Supervision. Tian Yin: Data curation. Haibing He: Funding acquisition. Yu Zhang: Validation. Xuemei Zhang: Project administration. Ke Liu: Methodology. Peifu Xiao: Writing – review & editing, Formal analysis. Xing Tang: Funding acquisition, Conceptualization. Declaration of competing interest The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper. Acknowledgments This work supported by National Key Research and Development Program of China [2024YFE0111300] and Hebei Province National Science Foundation [H2023109014]. Footnotes Supplementary data to this article can be found online at https://doi.org/10.1016/j.mtbio.2026.103654. Contributor Information Peifu Xiao, Email: peifuxiao@126.com. Xing Tang, Email: tanglab@126.com. Supplementary data The following is the Supplementary data to this article: Multimedia component 1 Data availability Data will be made available on request. References Associated Data Supplementary Materials Multimedia component 1 Data Availability Statement Data will be made available on request.