{"paper_id":"bcc841cf-ced3-472a-a2bd-4e2794b2314e","body_text":"Polydatin, also known as piceid, is a naturally occurring stilbenoid glucoside and a characteristic constituent of  Reynoutria japonica  (syn.  Polygonum cuspidatum ) [ 1 ]. As a purified compound with defined composition, polydatin permits controlled dosing and clearer attribution of pharmacological and formulation effects than crude botanical extracts [ 2 , 3 ]. Pharmacological studies have examined its effects in cardiovascular and cerebrovascular injury, metabolic and renal disorders, inflammatory and fibrotic diseases, and tissue repair [ 4 , 5 , 6 , 7 ]. Across these settings, polydatin has been associated with redox regulation, mitochondrial and endoplasmic reticulum homeostasis, innate immune and inflammasome signaling, regulated cell death, autophagy, and mitophagy [ 8 , 9 , 10 , 11 , 12 ]. Many of these mechanisms operate within specific tissues, cell populations, and intracellular compartments, making tissue, cellular, and intracellular exposure relevant to the magnitude and duration of the observed pharmacological responses.\nAchieving exposure sufficient to engage these mechanisms depends on the route of administration. Oral dosing is influenced by dissolution, intestinal transport, hydrolysis, phase-II metabolism, and systemic clearance, which contribute to low bioavailability and variable systemic exposure [ 13 , 14 , 15 , 16 ]. Intravenous administration avoids gastrointestinal processing but remains subject to stability in circulation, mononuclear phagocyte system clearance, tissue distribution, cellular uptake, and intracellular release [ 17 , 18 ]. Local administration instead depends on residence at the application site, tissue penetration, release duration, carrier degradation, and local tolerance [ 19 , 20 ]. Although the dominant barrier differs by route, each can limit the concentration or duration of polydatin exposure at relevant sites of action.\nPolydatin delivery research now spans oral, injectable, and local formulations designed to address route-specific limitations in exposure and tissue residence [ 21 , 22 , 23 , 24 , 25 ]. Unlike earlier reviews centered on general pharmacology, pharmacokinetics, or selected disease areas [ 4 , 6 ], this review asks whether formulation strategies change exposure, distribution, or efficacy and distinguishes nanocarriers from non-nano enabling systems. It thereby defines which approaches have demonstrated delivery value and which still lack the pharmacokinetic, safety, quality, or clinical evidence required for translation ( Figure 1 ).\nThis critical narrative review was informed by a structured PubMed search conducted on 7 July 2026 and updated on 26 August 2026. Search terms combined polydatin/piceid with concepts related to delivery systems, formulation and release, solid-state and local-delivery approaches, pharmacokinetics, bioavailability, biodistribution, analytical methods, and delivery-relevant mechanisms. The searches retrieved 590 record occurrences; after removal of 156 duplicate occurrences, 434 unique records underwent title/abstract screening. Original experimental studies were eligible when polydatin/piceid was a defined payload or formulation component and they reported delivery, formulation, exposure, distribution, or prespecified mechanistic data. Eligible non-nano formulation strategies were included as contextual comparators and analyzed separately from nanocarriers. Records without a defined polydatin intervention or relevant endpoint, carrier-only or experimentally unvalidated computational studies, non-primary reports, corrections, retractions, and botanical or food-chemistry records unrelated to delivery were excluded. Screening retained 157 records: 38 core delivery/formulation, 31 supporting formulation/pharmacokinetic, and 88 mechanistic/background records. Complete search strings, eligibility criteria, study-level classifications [ 26 , 27 ], exclusion reasons, and the adapted PRISMA-style flow are provided in  Supplementary Tables S1–S4 and Figure S1 ; the complete study-level mechanistic evidence matrix is provided in  Supplementary Table S5 .\n\nPolydatin, also known as piceid, is the 3-O-β-D-glucopyranoside of resveratrol. It has a molecular formula of C 20 H 22 O 8  and a molecular weight of 390.38 g/mol and consists of a stilbene scaffold with a β-D-glucopyranosyl residue linked through the phenolic hydroxyl group at C-3 ( Figure 2 ) [ 15 , 28 ]. It is a characteristic stilbenoid constituent of the rhizomes and roots of  Reynoutria japonica  Houtt. [syn.  Polygonum cuspidatum  Siebold & Zucc.;  Fallopia japonica  (Houtt.) Ronse Decr.] and has also been detected in grapes, red wine, peanuts, hop cones or pellets, and cocoa- or chocolate-containing products [ 1 , 15 ]. Both trans- and cis-isomers exist, although trans-polydatin is the form most commonly investigated. The configuration is analytically relevant because irradiation of trans-polydatin in solution can induce conversion to the cis-isomer [ 29 ].\nThe glucose residue changes several properties of the stilbene structure. Relative to resveratrol, polydatin has greater molecular size, polarity, and hydrogen-bonding capacity, together with different crystal packing, aqueous dissolution, membrane interaction, and susceptibility to transporter- and glycosidase-mediated processes. These differences affect absorption and metabolism and preclude treating polydatin and resveratrol as interchangeable compounds. Polydatin is therefore better regarded as a distinct stilbenoid glucoside whose biological disposition includes, but is not limited to, conversion to resveratrol [ 13 , 28 ].\nThe glycosidic bond is susceptible to microbial and enzymatic hydrolysis. Fermentation of  P. cuspidatum  with  Aspergillus oryzae  increased resveratrol formation through conversion of piceid, while cell extracts from several  Bifidobacterium  and  Lactobacillus  strains also catalyzed deglycosylation [ 30 , 31 ]. Although these experiments used fermentation or cell-extract systems, they establish that polydatin can serve as a substrate for microbial β-glucosidase activity. Intestinal epithelial studies further show that trans-piceid can undergo transport and deglycosylation concurrently, followed by glucuronidation of the resulting resveratrol [ 13 ]. Consequently, biological samples obtained after polydatin administration may contain intact polydatin, resveratrol, and conjugated forms of both compounds, each with potentially different distribution and biological activity.\nThe physicochemical limitations of polydatin cannot be described simply as generalized chemical instability. In a preformulation study, crystalline trans-piceid showed negligible degradation under accelerated heat and humidity, ambient fluorescent light, and International Council for Harmonisation photostability conditions; the chromatographic main-peak area remained between 98.04% and 99.35% across the 12-week study [ 32 ]. Solid-state form nevertheless affects pharmaceutical behavior. An anhydrous polymorph, a trihydrate, and an amorphous form have been characterized; the two crystalline forms remained stable during the reported storage study, whereas the amorphous material converted rapidly to the anhydrous form. After an equal oral dose in rats, the trihydrate showed faster absorption and approximately 1.5-fold greater exposure than the anhydrous polymorph [ 28 ]. Together with the solution-state photoisomerization described above, these data indicate that stability and exposure depend on physical form and experimental conditions rather than on a single intrinsic property.\nAt the intestinal epithelium, transport occurs alongside metabolism. In Caco-2 monolayers, trans-piceid displayed measurable bidirectional permeability in the 10 −6  cm/s range, but the apparent permeability decreased during six hours of exposure as resveratrol and glucuronide metabolites appeared [ 13 ]. Experiments with rat intestinal fractions implicated membrane-associated lactase-phlorizin hydrolase and cytosolic β-glucosidase in deglycosylation, with D-gluconolactone inhibiting both activities. The authors proposed two routes: luminal hydrolysis by lactase-phlorizin hydrolase followed by passive entry of resveratrol, and uptake of intact trans-piceid followed by cytosolic hydrolysis. Resveratrol was subsequently converted predominantly to resveratrol-3-O-β-D-glucuronide and, to a lesser extent, resveratrol-4′-O-β-D-glucuronide, with UGT1A1 implicated in this process [ 13 ]. Oral absorption therefore comprises several concurrent processes rather than transport of intact polydatin alone. Analyses of intestinal contents and urinary metabolites further identified resveratrol, dihydroresveratrol, sulfate conjugates, and glucuronide products after polydatin administration, supporting contributions from intestinal conversion and phase-II metabolism to its disposition [ 33 , 34 ].\nRat pharmacokinetic studies are consistent with extensive presystemic conversion and limited systemic availability. After gavage at 50, 100, or 300 mg/kg, polydatin, resveratrol, glucuronidated polydatin, and glucuronidated resveratrol were detected in plasma within 10 min; intestinal and hepatic perfusion experiments supported first-pass deglycosylation and glucuronidation [ 35 ]. In a separate LC-MS/MS study, the same range of oral doses produced substantially lower dose-normalized exposure than a 20 mg/kg intravenous reference, giving an absolute oral bioavailability of approximately 2.9% [ 14 ]. At an oral dose of 20 mg/kg, simultaneous quantification showed a Tmax of approximately 0.33 h for both polydatin and resveratrol and a polydatin terminal half-life of approximately 1 h. Rapid hydrolysis by rat fecal S9 fractions provided additional evidence for microbial contribution to resveratrol formation [ 15 ]. A subsequent disposition study found that trans-resveratrol-3-O-glucuronide reached higher concentrations than intact polydatin and resveratrol in plasma, urine, and bile and accounted for 52.8% of the administered dose recovered in urine over 72 h [ 36 ].\nHuman evidence is restricted to exploratory bioanalysis. In six volunteers receiving commercial tablets containing 40 or 75 mg polydatin, plasma polydatin was quantifiable in only two samples, at 10.06 and 65.07 nM, and urinary polydatin was detected in one participant after repeated intake [ 16 ]. Sparse sampling and the lower analytical sensitivity for resveratrol prevent this study from defining human pharmacokinetics or the complete metabolite profile, but the observations are consistent with low and variable circulating concentrations of polydatin.\nCurrent evidence therefore characterizes orally administered polydatin by physical-form-dependent behavior, measurable intestinal permeability, extensive intestinal and hepatic conversion, and low systemic availability. Plasma exposure has been examined more thoroughly than tissue-specific distribution, while sulfate and glucuronide profiles in target organs remain insufficiently defined. Further clarification of these exposure relationships will strengthen the mechanistic interpretation of polydatin studies and provide a pharmacokinetic basis for evaluating formulation-dependent changes in bioavailability and tissue distribution.\n\nMechanistic studies link polydatin to redox regulation, inflammatory signaling, organelle homeostasis, and regulated cell death [ 8 ]. Most evidence comes from cells and animal models treated with free polydatin. Target-binding assays, pathway inhibitors, and genetic interventions provide stronger mechanistic support than concurrent changes in signaling proteins and disease-related endpoints ( Table 1 ,  Figure 3 ).  Table 1  summarizes the mechanistic categories most relevant to delivery design; the complete study-level evidence matrix is provided in  Supplementary Table S5 .\nRedox regulation is consistently reported for polydatin. In cell-free micellar and liposomal systems, piceid inhibited linoleic acid peroxidation, indicating protection against membrane lipid peroxidation in addition to solution-phase radical scavenging [ 37 ]. In cultured cardiomyocytes and rats exposed to angiotensin II (Ang II), polydatin reduced NADPH oxidase activity, NOX2/NOX4 expression, and superoxide generation, changes accompanied by less hypertrophic remodeling [ 38 ]. Similar coupling of redox, inflammatory, and apoptotic responses was observed in D-galactose-induced liver and brain injury [ 39 ].\nPolydatin may also limit oxidative injury by preserving mitochondrial integrity. In rats subjected to hemorrhagic shock, polydatin preserved mitochondrial membrane potential in arterial smooth muscle cells, limited mitochondrial permeability-transition-pore opening and swelling, and reduced lysosomal destabilization [ 40 ]. In IL-1β-stimulated chondrocytes and osteoarthritic joints, polydatin increased SIRT3 and SOD2 expression, lowered mitochondrial superoxide production, and improved mitochondrial membrane potential, supporting SIRT3/SOD2-associated mitochondrial regulation [ 41 ].\nThe Keap1/Nrf2 pathway connects redox regulation with inflammatory signaling. In fructose-exposed hepatic cells and rats, polydatin increased microRNA-200a, reduced Keap1 expression, enhanced Nrf2-dependent antioxidant responses, and suppressed ROS-sensitive TXNIP/NLRP3 activation [ 42 ]. A microRNA-200a mimic and Keap1 or TXNIP silencing linked Keap1 suppression and Nrf2 activation to reduced oxidative stress and inflammasome signaling. Collectively, these studies describe polydatin as a regulator of redox-sensitive signaling and organelle function rather than a nonspecific antioxidant.\nPolydatin acts at several levels of inflammatory signaling. In mast cells, IgE/antigen-mediated FcεRI crosslinking activates Lyn and Syk, followed by MAPK, PI3K/Akt, NF-κB, calcium mobilization, and mediator release. Polydatin reduced Lyn and Syk activity in cell-free kinase assays, attenuated downstream signaling, and increased Nrf2/HO-1 signaling, with corresponding reductions in degranulation and cytokine production [ 43 ]. The kinase assays place the effect of polydatin at receptor-proximal signaling in addition to changes in downstream cytokine production.\nMembrane organization and sirtuin signaling provide additional regulatory sites. In LPS-stimulated BV2 microglia, polydatin decreased cholesterol in isolated lipid-raft fractions and reduced lipid-raft formation, coupled to lower PI3K/Akt/NF-κB activation and inflammatory mediator production [ 44 ]. In sepsis-associated encephalopathy, polydatin increased SIRT1 activity and reduced p38 phosphorylation, mitochondrial dysfunction, ROS generation, and inflammatory cytokines. The SIRT1 inhibitor Ex527 attenuated these effects, whereas p38 inhibition reproduced part of the response, supporting the involvement of SIRT1-regulated p38 signaling [ 45 ].\nPolydatin also modifies immune-cell effector functions. In lupus-prone mice and neutrophils from patients with systemic lupus erythematosus, polydatin reduced ROS production, neutrophil extracellular trap formation, and associated autoimmune pathology [ 46 ]. In crystal-induced inflammatory models, polydatin reduced IL-1β-related responses in THP-1 cells and was associated with PPAR-γ induction, lower NLRP3 signaling, and changes in ferritin-related iron handling in monosodium urate-induced arthritis [ 47 , 48 ]. In osteoarthritis models, polydatin also modified macrophage polarization and bone-remodeling responses [ 49 ]. No single inflammatory target is shared across these models. Polydatin instead affects receptor-proximal kinases, lipid-raft-associated signaling, redox-sensitive transcriptional pathways, and immune-cell effector responses in a cell- and stimulus-dependent manner.\nNLRP3 inflammasome activation integrates ionic disturbance, organelle stress, lysosomal injury, and endogenous damage signals. Following priming and activation, assembly of the NLRP3-ASC-caspase-1 complex promotes IL-1β and IL-18 maturation and gasdermin D-mediated pyroptosis [ 50 ]. Polydatin has been reported to affect inflammasome priming, complex stability, and pyroptotic execution.\nIn potassium oxonate-induced hyperuricemia and renal inflammation, polydatin reduced IκBα degradation, NF-κB p65 activation, and NLRP3/ASC/caspase-1 signaling while increasing AMPK phosphorylation and SIRT1 expression [ 51 ]. In dry eye models, reduced ROS generation and NF-κB nuclear translocation were accompanied by lower NLRP3 expression and caspase-1 activation in corneal epithelial cells [ 52 ]. Polydatin also reduced NLRP3/IL-1β/NF-κB signaling, mesothelial inflammatory responses, and fibrotic changes in high-glucose-induced peritoneal fibrosis [ 53 ]. These studies support inhibition of redox- and NF-κB-related inflammasome priming, although neither identifies a direct molecular target shared across models.\nMore specific target evidence has been reported for HSP90α. Lip-SMap protein profiling and cellular thermal shift assays identified HSP90α as a polydatin-binding protein. Microscale thermophoresis and isothermal titration calorimetry yielded dissociation constants of 1.37 and 1.34 μM, respectively, and polydatin inhibited HSP90α ATPase activity with an IC50 of approximately 2.73 μM. HSP90α deletion abolished further suppression of IL-1β secretion by polydatin, while co-immunoprecipitation and imaging experiments showed disruption of the HSP90α-NLRP3 interaction and loss of resting and activated NLRP3 complexes [ 11 ]. The combined binding, target-engagement, and genetic evidence supports HSP90α as a direct molecular target in this experimental setting. In gouty nephropathy, polydatin also reduced caspase-1 activation, GSDMD and GSDMD-N expression, IL-1β release, LDH release, and renal tubular epithelial-cell death, linking inflammasome inhibition with reduced pyroptotic membrane permeabilization [ 54 ].\nFerroptosis is an iron-dependent form of regulated cell death driven by membrane lipid peroxidation, with GPX4-mediated lipid hydroperoxide reduction serving as a central defense mechanism [ 55 ]. In hemin-treated Neuro2A cells and a traumatic brain injury model, polydatin reduced iron accumulation and membrane lipid peroxidation and restored selenium-dependent glutathione peroxidase activity, accompanied by changes in GPX4- and SLC7A11-related responses [ 56 ]. In pancreatic β-cells exposed to palmitate under high-glucose conditions, molecular docking and a GPX4 promoter-luciferase reporter screen identified polydatin as a candidate regulator of GPX4 expression. Polydatin increased GPX4, SLC7A11, and Nrf2-related responses and reduced ferroptosis-associated changes [ 12 ]. The reporter system measured GPX4 promoter regulation rather than direct GPX4 enzymatic activation; these findings therefore support regulation of the Nrf2/GPX4 defense system rather than direct binding to GPX4. Overall, polydatin affects inflammasome-dependent pyroptosis and ferroptosis through distinct molecular processes.\nPolydatin has been reported to regulate autophagy and selective mitochondrial clearance in a context-dependent manner; functional perturbation, rather than marker changes alone, is needed to relate these processes to tissue protection [ 57 , 58 ]. In septic myocardial injury, polydatin increased SIRT6 expression and produced an autophagy-associated profile characterized by increased LC3 processing and Beclin-1 and reduced p62. SIRT6 silencing attenuated these marker changes, and 3-methyladenine weakened the reductions in apoptosis and inflammatory cytokines, supporting a functional contribution of SIRT6-dependent autophagy [ 59 ]. In atherosclerotic mice, pharmacological inhibition or induction of autophagy respectively weakened or reinforced the response to polydatin, linking PI3K/Akt/mTOR inhibition and autophagy restoration to reduced plaque formation [ 60 ]. Mitophagy-related evidence includes human trisomy 21 fibroblasts, in which polydatin reduced microRNA-155, restored BAG5, PINK1, and CBL expression, increased mitochondrial-lysosomal colocalization, and improved mitochondrial bioenergetics [ 61 ]. A later study in fatty-acid-exposed HepG2 cells and high-fat-diet rats linked the effects of free polydatin on mitochondrial dynamics and PINK1/Parkin-associated mitophagy to BMAL1; BMAL1 knockdown abolished the reported metabolic and mitochondrial responses [ 62 ]. Both studies used free polydatin and therefore provide mechanistic endpoints rather than delivery-system evidence.\nSeveral mechanistic features of polydatin impose specific requirements on delivery ( Table 1 ). Nrf2/Keap1, HSP90α/NLRP3, GPX4, sirtuins, and the autophagy machinery are intracellular, while mitochondrial and lysosomal effects require access to particular subcellular environments [ 11 , 42 , 59 ]. The relevant cell population also varies among disease settings, including inflammatory cells, vascular and epithelial cells, chondrocytes, and neurons. Tissue accumulation alone does not establish that polydatin has reached the cells or intracellular sites required for pathway modulation [ 63 , 64 ]. The available studies do not define a common concentration threshold or exposure duration. Acute injury and chronic inflammatory models use different schedules, but few compare rapid with sustained exposure. Cell studies relate nominal extracellular concentrations to pathway changes without measuring intracellular polydatin or concentrations required for target engagement. Delivery studies could address these gaps by linking drug release, cellular uptake, and intracellular exposure to pathway-specific endpoints and to the magnitude and duration of local exposure.\n\nBecause the same carrier can perform differently after oral, intravenous, intraperitoneal, or local administration, this section is organized primarily by administration route rather than material class. Evidence is considered at three levels: formulation properties, including solubility, stability, and release; delivery performance, including pharmacokinetics, tissue distribution, and cellular uptake; and biological responses in experimental models. Improvement at one level does not establish improvement at the others ( Figure 4 ).\nOral formulations pursue two distinct goals: increasing systemic exposure or prolonging intestinal residence. Non-nano strategies provide important comparators. Cyclodextrin inclusion complexes [ 29 , 65 ] and an L-proline co-crystal [ 66 ] improved in vitro solubility or stability but were not linked to oral pharmacokinetics. By contrast, a trihydrate polymorph produced faster absorption and approximately 1.5-fold greater exposure than the anhydrous form after a matched oral dose [ 28 ]. Among nanoscale systems, oral liposomes provided the first direct pharmacokinetic comparison with a dose-matched polydatin suspension [ 67 ]. The formulation increased AUC 2.83-fold and prolonged apparent half-life and mean residence time, whereas the 1.21-fold increase in Cmax was not reported as significant. A separate ultrasound-assisted liposome study optimized encapsulation and processing but did not evaluate pharmacokinetics or in vivo delivery [ 68 ].\nPolymeric and polysaccharide carriers have used chitosan, PVP, and PLGA to improve wetting, mucoadhesion, or release control [ 82 , 83 ]. Dose-matched formulations generally produced stronger pharmacodynamic responses than free polydatin in diabetes [ 21 ], associated nephropathy and liver injury [ 84 , 85 ], and oxaliplatin-induced intestinal injury or developing pulmonary fibrosis [ 69 , 70 ]. However, these studies did not measure pharmacokinetics, mucus residence, tight-junction modulation, or epithelial transport, so the basis of the improved oral responses remains unresolved.\nTwo sequential studies from the same group evaluated fucoidan-based oral nanoparticles in cisplatin-induced kidney injury. The first used a fucoidan/carboxymethyl-chitosan carrier and reported a 2.33-fold increase in AUC together with higher renal polydatin concentrations than free drug [ 71 ]. The later formulation incorporated the pharmacologically active chemical chaperone 4-PBA and reported an approximately twofold AUC increase together with P-selectin-associated uptake and endoplasmic reticulum localization experiments [ 72 ]. Because both the carrier and 4-PBA may contribute to the response, the later report is best interpreted as an extension of the earlier platform rather than independent replication. Across the three direct oral comparisons, formulated polydatin increased AUC by approximately 2.0–2.83-fold, whereas changes in Cmax ranged from 1.21- to 3.77-fold. The two fucoidan studies were small ( n  = 3 per group) and contained internal inconsistencies in the reported AUC units;  Table 2  therefore reproduces the source values. Collectively, these data support increased systemic exposure but not selective kidney delivery, and they do not resolve intestinal conversion or metabolite disposition.\nOther oral systems are intended to increase intestinal exposure rather than systemic absorption. PVP-polydatin nanoparticles improved outcomes in oxaliplatin-induced intestinal injury, but intestinal polydatin concentrations were not measured [ 70 ]. A graphene oxide-polyacrylic acid hydrogel was designed to swell after oral administration and sustain polydatin release during radiation injury [ 73 ]. For these systems, gastrointestinal residence and polydatin concentrations in mucus, epithelium, and deeper tissue are more informative than plasma exposure alone.  Table 2  and  Table 3  summarize direct PK and broader oral evidence.\nThe existing injectable studies used either intravenous or intraperitoneal administration. Both avoid gastrointestinal processing, but the delivery questions are not identical. Intravenous carriers require evaluation of interactions with blood components, mononuclear phagocyte system clearance, vascular transport, tissue distribution, and cellular internalization [ 17 , 18 ]. Intraperitoneal administration also avoids intestinal absorption but should not be treated as pharmacokinetically equivalent to intravenous injection ( Table 4 ).\nThe principal intravenous study used ROS- and pH-responsive micelles in carbon tetrachloride-induced liver fibrosis. After 3 weeks of model induction, mice received the micelles by tail-vein injection every other day for a further 3 weeks. Each dose contained 40 mg/kg carrier and 2.5 mg/kg polydatin; blank micelles were given intravenously at 40 mg/kg, whereas free polydatin was administered intraperitoneally at 2.5 mg/kg [ 23 ]. The micelles accumulated in the liver, with early uptake attributed largely to the hepatic mononuclear phagocyte system. Fluorescence imaging indicated greater probe release in fibrotic liver, but measured organ polydatin concentrations did not differ between healthy and fibrotic mice. The route mismatch between micellar and free polydatin groups limits direct comparison of efficacy.\nOther injectable studies used intraperitoneal administration. Exosome-coated polydatin nanoparticles and route-matched control preparations were injected immediately after abdominal irradiation; the methods reported 20 mg/kg polydatin and 2 mg/kg exosomes [ 22 ]. The formulation increased survival and altered intestinal histology and mitochondrial functional measurements, but comparative plasma or tissue concentration-time data were not reported. The study therefore supports biological activity after exosome-associated delivery but does not establish increased bioavailability. Polydatin-associated gold nanoparticles were administered intraperitoneally for 21 days in tumor-bearing mice, with a 20 mg/kg free-polydatin group and blank gold nanoparticle controls [ 74 ]. Although the route-matched controls aid interpretation, incomplete reporting of polydatin loading, release, and pharmacokinetics limits assessment of the carrier contribution. Injectable administration has therefore been shown to be feasible, but current evidence does not establish a consistent improvement in the systemic disposition or tissue delivery of polydatin.\nLocal delivery shifts the evaluation focus from plasma exposure to residence at the application site, tissue penetration, release duration, material degradation, and local tolerance [ 19 , 20 ]. Polydatin-loaded polycaprolactone nanofibers released polydatin for up to 14 days without a marked initial burst and supported osteogenic differentiation in vitro [ 77 ]. No implantation study was performed, so residence, degradation, and local polydatin exposure remain to be established in vivo.\nA collagen-oxidized pullulan scaffold released approximately 63% of its polydatin within 48 h and more than 90% within 96 h. When applied to excisional wounds, the polydatin-containing scaffold was compared with untreated wounds, collagen alone, and collagen-oxidized pullulan controls and was followed for up to 16 days in normal rats and 21 days in diabetic rats [ 25 ]. The matrix increased wound closure and altered tissue-repair markers, but local polydatin concentrations and scaffold degradation were not quantified. A more recent oxidized hyaluronic acid/catechol-chitosan hydrogel contained 0.4 mg/mL polydatin and 0.068 mg/mL tetramethylpyrazine and released approximately 80% of both agents within 48–52 h [ 78 ]. Wounds were followed for 14 days, but the in vivo comparison included the blank hydrogel and the dual-drug hydrogel without single-drug hydrogel groups; the individual contribution of polydatin cannot be determined. A carrier-free hydrogel co-assembled from mangiferin and polydatin has been reported for MRSA-infected diabetic wounds, with demonstrated sustained release, antibacterial and anti-biofilm effects, macrophage polarization, revascularization, and improved wound repair in rats. However, the absence of component-specific controls means these outcomes can only be attributed to the complete non-nano binary hydrogel, not to polydatin alone [ 91 ]. These local delivery systems and their principal evidence boundaries are summarized in  Table 5 .\nDirect mucosal delivery has been explored using formulations placed on oral surfaces. Extract-containing electrospun nanofibers improved the solubility and mucosal permeation of polydatin and resveratrol [ 92 ], whereas mucoadhesive tablets prolonged local release and mucosal residence [ 79 ], and perio-dontal films provided sustained local release [ 80 ]. A separate study used hot-melt extrusion to prepare solid dispersions of polydatin-rich  Polygoni cuspidati  extract with polyvinyl polymers and incorporated selected extrudates into mucoadhesive buccal tablets; the resulting formulation increased polydatin release and showed in vitro mucoadhesive performance [ 93 ]. Because these preparations contain multicomponent botanical extracts, their release and biological effects cannot be assigned to polydatin alone ( Table 6 ).\nTargeted and responsive systems seek to control complementary stages of polydatin delivery: targeted carriers alter tissue or cellular distribution, whereas responsive materials regulate drug release or carrier activity after exposure to a local stimulus. Evidence for targeting ranges from receptor-blocking experiments in cells to ligand-controlled distribution studies in vivo. In A549 cells, hyaluronic acid- and lactoferrin-coated polydatin/PLGA nanoparticles increased cellular uptake, and CD44 blockade reduced uptake of the hyaluronic acid-coated formulation [ 76 ]. The oral fucoidan/4-PBA system used P-selectin inhibition to examine carrier uptake and fluorescence colocalization to assess endoplasmic reticulum localization [ 72 ]. The intratracheally administered SPFPT-modified lipid nanoparticle included both ligand and component controls when delivering polydatin and siRNA targeting importin 7 to mechanically activated alveolar epithelial cells [ 75 ]. Importin 7 silencing reduced p38 nuclear translocation, whereas polydatin contributed to NF-κB inhibition. Cy5-labeled formulations containing 5 μg siRNA were assessed 2 h after administration against untargeted and scrambled-peptide controls, and the combined formulation reduced TNF-α, IL-6, and HMGB1 more than the polydatin-only or siRNA-only nanoparticles. These experiments associate specific ligands with carrier uptake or accumulation, although the receptor for SPFPT remains unidentified. Because fluorescence does not directly measure drug delivery, quantitative imaging combined with polydatin measurement would distinguish carrier biodistribution from target-cell drug exposure [ 94 ].\nTaken together, these examples illustrate three evidence levels that should not be conflated. Fluorescence-only findings indicate carrier or surrogate distribution, whereas CD44 blockade in A549 cells and SPFPT negative-control experiments support mechanism-based carrier targeting; however, the tracked analytes were RhB and Cy5-siRNA rather than polydatin [ 75 , 76 ]. The fucoidan systems measured tissue polydatin, but only at a single time point and across multiple organs. Moreover, mouse DiR renal fluorescence was not concordant with rat tissue-polydatin concentrations, which were highest in the liver for Fu-4-PBA/polydatin nanoparticles [ 71 , 72 ]. These limitations preclude a drug-level active kidney-targeting claim, and none of the current studies combines quantitative tissue polydatin with a matched non-targeted comparator and mechanistic evidence. This distinction provides the context for evaluating stimulus-responsive release separately from selective drug targeting.\nAfter a carrier reaches the intended tissue or cell population, stimulus-responsive design can regulate polydatin release or activate a functional carrier. Intravenous micelles released 9.38% of polydatin over 24 h at pH 7.4 and 89.27% under combined pH 5.0 and oxidative conditions [ 23 ]. Because pH 5.0 is more representative of endosomal or lysosomal compartments than the extracellular environment of fibrotic liver [ 95 ], this profile is more consistent with activation after cellular uptake. Two ZIF-8 systems illustrate a further distinction. A Cu-doped ZIF-8 nanozyme hydrogel combined pH-responsive carrier activity with local antibacterial and wound-healing effects [ 81 ], whereas chitosan-coated polydatin@ZIF-8 nanoparticles showed 91.34% loading efficiency, pH-dependent release, and in vitro antibacterial activity without an in vivo exposure, efficacy, or selective-targeting experiment [ 90 ]. These systems demonstrate responsive release or carrier activity, but neither acidic release nor fluorescence alone establishes selective delivery of polydatin ( Table 4  and  Table 5 ).\n\nFormulation characterization, release studies, pharmacokinetics, and cellular uptake describe delivery performance, whereas disease models allow the pharmacological relevance of a formulation to be assessed. In these models, the key question is whether altered release, exposure, or distribution changes the magnitude or duration of the pharmacological effects of polydatin in affected tissues and cells. Effects contributed by carrier materials or co-delivered agents must also be distinguished from those attributable to polydatin. Polydatin delivery systems have been examined in models of diabetes and its complications [ 21 ], kidney and intestinal injury [ 70 , 71 ], and pulmonary and hepatic fibrosis [ 23 , 69 ]. Cancer and tissue-repair applications have also been explored [ 24 , 25 ]. The following subsections examine the pharmacological effects reported in these models and the evidence linking them to polydatin delivery ( Table 7 ).\nOral chitosan nanoparticles have been examined in a closely related experimental series using nicotinamide/streptozotocin-induced type 2 diabetic rats. In the systemic metabolic study, polydatin-loaded chitosan nanoparticles and free polydatin were administered by oral gavage at 50 mg/kg once daily for 28 days. Compared with dose-matched free polydatin, the nanoparticle formulation produced larger reductions in fasting glucose, HbA1c, and insulin resistance, increased hepatic glycogen, and improved the serum lipid profile [ 21 ]. Studies using a similar formulation, dose, and four-week schedule extended these findings to early diabetic nephropathy and diabetic liver injury. The nephropathy study reported greater improvements in renal function indices, oxidative stress, NF-κB/COX-2 signaling, and circulating cytokines [ 85 ], whereas the liver study showed improved hepatic glucose metabolism and antioxidant status, lower transaminase activities, and modulation of GLUT2, glucokinase, TNF-α, and IL-1β [ 84 ]. The consistency of these metabolic, renal, and hepatic responses supports greater oral pharmacodynamic activity of the chitosan formulation than free polydatin, although the exposure changes responsible for this difference were not measured.\nLocal delivery has extended polydatin application to diabetic wounds and ulcers. In diabetic excision wounds, complete closure occurred by day 21 with a polydatin-containing collagen-oxidized pullulan scaffold, compared with day 27 for the unloaded scaffold and day 30 in untreated animals [ 25 ]. An oxidized hyaluronic acid/catechol-chitosan hydrogel co-delivering tetramethylpyrazine and polydatin produced nearly complete closure by day 14 and altered markers of angiogenesis and matrix remodeling [ 78 ]. An infection-responsive hydrogel containing polydatin-loaded Cu-doped ZIF-8 nanozymes reduced bacterial burden and accelerated closure of diabetic oral ulcers [ 81 ]. The scaffold provided sustained local polydatin delivery, whereas the two hydrogels combined polydatin with complementary proangiogenic or antibacterial components. The effects of the latter formulations therefore reflect the complete delivery systems rather than polydatin alone.\nDiabetic vascular and myocardial injury represents a further setting for polydatin delivery. Under high-glucose conditions, free polydatin restored endothelium-dependent relaxation through PPARβ/NO signaling and preserved endothelial mitochondrial membrane potential and Drp1-associated mitochondrial dynamics [ 96 , 97 ]. In diabetic myocardial models, Caveolin-1 deletion or knockdown attenuated the response to polydatin, while a separate vascular study reported lower NLRP3 and VCAM-1 expression after polydatin treatment [ 98 ]. These studies provide defined pharmacodynamic endpoints for evaluating polydatin delivery in diabetic vascular and myocardial injury.\nTwo sequential studies from the same group linked oral pharmacokinetics and organ distribution with renal outcomes in cisplatin nephrotoxicity. The first used self-assembled fucoidan/carboxymethyl-chitosan nanoparticles, which increased AUC approximately 2.3-fold and increased renal polydatin concentrations relative to free polydatin. The formulation also produced larger reductions in renal dysfunction, tubular injury, oxidative stress, DNA damage, cytosolic mitochondrial DNA, and cGAS-STING-related inflammation [ 71 ]. The later system incorporated 4-PBA into the fucoidan carrier and approximately doubled AUC while adding P-selectin-associated uptake and endoplasmic reticulum localization experiments. It also produced stronger renal and molecular responses than free polydatin [ 72 ]. However, 4-PBA is an active chemical chaperone, so these effects reflect both altered polydatin exposure and the pharmacological contribution of the co-component. Both studies began treatment before cisplatin administration and should therefore be interpreted as prevention or attenuation of developing injury. Their shared investigators, model, and endpoints also mean that the second study extends, rather than independently replicates, the first. Higher renal concentrations occurred alongside substantial hepatic distribution, supporting enhanced renal exposure but not kidney-selective delivery. Studies of free polydatin have additionally identified NF-κB inhibition and Nrf2/HO-1 activation, SIRT6-dependent autophagy, and reduced YAP expression or nuclear translocation as renal pharmacodynamic endpoints [ 99 , 100 , 101 ].\nIntestinal studies have examined systemic intervention after radiation injury, oral nanoparticle delivery during chemotherapy, and sustained luminal release. In the radiation-injury model, exosome-coated polydatin nanoparticles administered immediately after irradiation improved survival and leukocyte recovery and reduced diarrhea, villus loss, crypt injury, γ-H2AX, inflammatory mediators, mitochondrial dysfunction, and cellular senescence [ 22 ]. These findings indicate that post-injury administration supported both systemic recovery and intestinal epithelial preservation. In an oxaliplatin injury model, mice received free polydatin or polyvinylpyrrolidone-polydatin nanoparticles orally at 50 or 100 mg/kg for seven days, with oxaliplatin administered on days 3–6. At 100 mg/kg, the nanoparticle group showed greater preservation of body weight, colon length, crypt architecture, ileal villus morphology, goblet cells, and mucosal integrity than dose-matched free polydatin. Lower ROS, γ-H2AX, cGAS, STING, IL-1β, and TNF-α and improved mitochondrial indices linked the stronger tissue response to attenuation of epithelial stress and inflammation [ 70 ]. An orally administered graphene oxide-polyacrylic acid hydrogel was also reported to sustain intestinal polydatin release and improve epithelial and hematological endpoints after irradiation [ 73 ]. Free polydatin has separately been reported to preserve epithelial junction proteins and reduce Nrf2/SLC7A11/GPX4-related ferroptotic injury in DSS colitis [ 102 ], providing an additional mechanism that could be examined in future intestinal delivery studies. Collectively, systemic rescue, oral nanoparticle delivery, and sustained luminal release have each extended the intestinal effects of polydatin across radiation-, chemotherapy-, and inflammation-related injury models.\nPulmonary delivery studies address two route-specific objectives: local intervention after acute mechanical injury and repeated oral administration during fibrosis development. In ventilator-induced lung injury, SPFPT-modified lipid nanoparticles co-delivering polydatin and siRNA targeting importin 7 were administered intratracheally after 4 h of mechanical ventilation. Compared with nanoparticles carrying either component alone, the combined formulation produced greater inhibition of p38 and NF-κB signaling, lowered TNF-α, IL-6, and HMGB1 in bronchoalveolar lavage fluid, and reduced pulmonary edema, neutrophil infiltration, and histological injury [ 75 ]. Cy5 imaging localized the siRNA-containing formulation to mechanically activated alveolar epithelial cells but did not measure the distribution of polydatin.\nBy contrast, an oral chitooligosaccharide-coated PLGA formulation was evaluated during the development of bleomycin-induced pulmonary fibrosis. The formulation was administered once daily at a dose equivalent to 25 mg/kg polydatin from day 1 to day 17; bleomycin was given 1 h after the third dose on day 3. Compared with dose-matched free polydatin, the nanoparticle group showed greater preservation of lung architecture and larger reductions in collagen deposition, α-SMA, TGF-β, lipid peroxidation, TNF-α, and IL-1β [ 69 ]. Because dosing preceded bleomycin exposure, the results demonstrate prevention or attenuation of developing fibrosis rather than intervention in established fibrosis.\nMechanistic studies of free polydatin extend the pulmonary evidence beyond these delivery models. After  Mycoplasma pneumoniae  infection, polydatin suppressed NLRP3/NF-κB signaling and caspase-1/GSDMD-related epithelial pyroptosis [ 103 , 104 ]. Other studies reported attenuation of methotrexate-induced pulmonary fibrosis and reduced S100B-dependent neutrophil extracellular trap formation in traumatic brain injury-associated lung injury [ 105 , 106 ]. These pathways provide defined pharmacodynamic endpoints for pulmonary delivery studies. Current formulation evidence therefore supports local co-delivery after acute mechanical injury and repeated oral administration during fibrosis development, with established pulmonary fibrosis representing a relevant setting for subsequent evaluation.\nROS- and pH-responsive micelles were evaluated in a carbon tetrachloride (CCl 4 )-induced liver fibrosis model using an intervention schedule initiated after fibrosis had begun to develop. Following three weeks of CCl 4  exposure, the micelles were administered by tail-vein injection every other day for a further three weeks at 40 mg/kg carrier, equivalent to 2.5 mg/kg polydatin. The formulation reduced hepatocyte apoptosis, inflammatory macrophage activity, hepatic stellate cell activation, collagen deposition, and fibrosis-related markers [ 23 ]. Cellular and tissue experiments were consistent with uptake by hepatocytes, macrophages, and hepatic stellate cells and with release under ROS-rich and acidic conditions. The micellar system integrates microenvironment-responsive release with pharmacological effects in several cell populations involved in fibrosis. The unloaded micelles also reduced oxidative stress, indicating that ROS scavenging by the carrier complemented the activity of polydatin. Thus, the reported antifibrotic response reflects the combined contribution of responsive delivery, polydatin, and the carrier material. Direct comparison with free polydatin remains limited by the use of intravenous micelles and intraperitoneal free polydatin.\nCancer-related formulations have been examined in chemoprevention, combination treatment, and targeted cellular delivery. Oral PLGA-polydatin nanoparticles administered at 7.5–30 mg/kg three times weekly throughout a 14-week 7,12-dimethylbenz[a]anthracene (DMBA) protocol reduced tumor incidence, tumor burden, lipid peroxidation, and proliferation-related markers in buccal carcinogenesis [ 24 ]. However, the absence of free-polydatin and unloaded-PLGA groups prevents separation of the formulation effect. In Ehrlich ascites carcinoma-bearing mice, intraperitoneal polydatin-gold nanoparticles containing 20 mg/kg polydatin, administered alone or with doxorubicin for 21 days, improved tumor-related and biochemical endpoints [ 74 ]. Hyaluronic acid/lactoferrin-coated PLGA nanoparticles increased CD44-associated uptake and cytotoxicity in A549 lung cancer cells [ 76 ], providing in vitro support for receptor-associated cellular delivery. In HER2-overexpressing SKBR3 cells, polydatin-loaded chitosan nanocapsules at 200 and 400 μg/mL reduced proliferative capacity by approximately 40% and 60%, respectively, after 72 h and produced cytoskeletal and nuclear changes, whereas empty nanocapsules did not reduce viability [ 88 ]. Gelatin-coated polycaprolactone nanoparticles produced stronger antiproliferative responses than free polydatin in radioresistant A549 cells and three-dimensional spheroids, although proteomic and in silico analyses indicated broadly similar molecular responses [ 89 ].\nMechanistic studies provide additional endpoints for cancer-oriented delivery. Combination studies reported that polydatin enhanced cisplatin-associated responses in non-small-cell lung cancer through NOX5-dependent ROS generation and endoplasmic reticulum stress/JNK/p38 signaling. During sunitinib treatment, polydatin reduced oxidative and inflammatory responses associated with NLRP3, MyD88, and NF-κB signaling. Separate work in head and neck squamous cell carcinoma linked polydatin to HDAC1 inhibition and reduced angiogenic activity [ 107 , 108 , 109 ]. Outside cancer, polymeric nanocapsules retained the anti-inflammatory and cytoprotective activity of polydatin in LPS-exposed hippocampal organotypic cultures [ 87 ], while amphiphilic chitosan nanocarriers improved behavioral, inflammatory, and oxidative endpoints in an aluminum chloride-induced Alzheimer-like rat model [ 86 ]. Polydatin-loaded polycaprolactone nanofibers promoted osteogenic differentiation in vitro [ 77 ], supporting their further evaluation in bone-repair models. Atherosclerosis represents another mechanism-based application because free polydatin has been reported to activate autophagy and suppress NLRP3 inflammasome-related pyroptosis in atherosclerotic lesions [ 110 ].\nAcross the disease models reviewed, dose-matched comparisons show that polydatin delivery systems produced larger pharmacodynamic responses than free polydatin in diabetes, renal and intestinal injury, and developing pulmonary fibrosis [ 21 , 69 , 70 , 71 ]. In cisplatin nephrotoxicity, two fucoidan studies combined plasma pharmacokinetics with single-time tissue measurements; stronger renal responses accompanied higher systemic exposure, but multi-organ enrichment did not establish kidney-selective delivery [ 71 , 72 ]. Local and responsive formulations prolonged release or residence and enabled stimulus-dependent material behavior [ 23 , 25 ], while multicomponent systems added antibacterial or combination functions [ 74 , 81 ]. Responses to the latter systems reflect the complete formulations rather than polydatin alone.\nSeveral mechanisms established for free polydatin provide additional pharmacodynamic endpoints for delivery studies. These include HSP90α/NLRP3 regulation and pyroptosis [ 11 , 54 ], Nrf2/system Xc − /GPX4-dependent ferroptosis control [ 12 , 56 , 111 ], SIRT-dependent autophagy and mitophagy [ 10 , 59 , 61 ], and immune-cell processes involving neutrophil extracellular traps and lipid-raft-associated signaling [ 44 , 46 ]. Incorporating these intracellular endpoints alongside quantitative exposure measurements would clarify whether formulation-dependent changes in polydatin disposition modify target engagement and regulated cell-death pathways in relevant cell populations.\n\nCurrent safety evidence consists mainly of short-term cytocompatibility, hemocompatibility, and selected animal observations. ROS- and pH-responsive micelles maintained more than 95% viability in hepatocytes, macrophages, and hepatic stellate cells at a polydatin concentration of 25 μg/mL [ 23 ]; a collagen-oxidized pullulan scaffold supported approximately 93% fibroblast viability and was also evaluated for hemolysis and erythrocyte adhesion [ 25 ]. Fucoidan/4-PBA nanoparticles produced a hemolysis rate of 4.92% at 10 μg/mL, and their cytocompatibility was examined in HK-2 cells [ 72 ]. In vivo, no overt histological injury was detected in the heart, liver, spleen, or lung after oral administration of fucoidan/carboxymethyl-chitosan nanoparticles at 100 mg/kg in the cisplatin nephrotoxicity study, while major-organ histology remained unremarkable 21 days after intratracheal administration of SPFPT-modified lipid nanoparticles [ 71 , 75 ]. These findings support only initial compatibility under the tested conditions; they do not establish repeat-dose safety, immune compatibility, or carrier clearance.\nAccordingly, a lead polydatin formulation should undergo route- and duration-matched repeat-dose studies with toxicokinetic confirmation, recovery cohorts, clinical observations, hematology and coagulation testing, serum chemistry, urinalysis, organ weights, histopathology, and assessment of carrier or excipient persistence. Empty-carrier and free-polydatin comparators are essential, particularly for fucoidan-, chitosan-, exosome-, gold-, siRNA-, or 4-PBA-containing systems. Immunogenicity and immunotoxicity assessment should be product-specific: blood-contacting or inhaled particles require evaluation of complement activation, cytokine release, leukocyte and platelet effects, hemolysis, and hypersensitivity risk; proteinaceous, extracellular-vesicle, or surface-ligand components may additionally require anti-carrier antibody testing where biologically plausible. Local scaffolds require evaluation of degradation products, irritation, delayed inflammation, and tissue remodeling. Their omission from the included studies is an evidence gap, not evidence of safety [ 112 ].\nExisting quality-control studies cover chemical identity, drug content, particle properties, stability, and route-relevant release. A validated fused-core HPLC method separated polydatin and resveratrol within 3 min and showed suitable precision and recovery for identity and content testing in commercial preparations [ 113 ]. Representative nanoliposomes had a mean particle size of 80.2 nm and an encapsulation efficiency of 88.4% [ 67 ], whereas oral chitosan nanoparticles measured 144.25 ± 3.37 nm, achieved an encapsulation efficiency of 96.74 ± 0.39%, and released less than 20% of polydatin within 12 h [ 21 ]. Fucoidan/4-PBA nanoparticles measured 102 ± 3.46 nm, had a zeta potential of −16.5 ± 0.49 mV and a drug-loading capacity of 10.34 ± 0.60%, and were monitored for particle-size and surface-charge stability over two months [ 72 ].\nRelease testing also reflected the intended formulation function: responsive micelles released 9.38% of polydatin over 24 h at pH 7.4 but 89.27% under combined acidic and oxidative conditions [ 23 ], whereas the collagen-oxidized pullulan scaffold released 62.6% within 48 h and more than 90% within 96 h [ 25 ]. These reports show that formulation-level measurements are feasible, but most are single-batch characterizations and do not establish manufacturing consistency or a product-level control strategy [ 27 , 114 ].\nConsistent with quality-by-design principles, a development-ready control strategy for polydatin should link drug-substance and drug-product attributes to clinically relevant performance. Drug-substance testing should distinguish trans-polydatin from cis-polydatin and resveratrol and address assay, water content, solid form, source-related impurities, residual solvents, and degradants formed by light, oxidation, hydrolysis, or deglycosylation. Drug-product assays should distinguish total polydatin from encapsulated, surface-associated, and unencapsulated fractions. Mean particle size alone is insufficient; distribution width or polydispersity index, particle concentration, morphology, surface charge, ligand density, carrier molecular weight, loading, in vitro release, pH, osmolality, and residual solvents or crosslinkers should be measured as relevant. Sterility, endotoxin, and container-closure integrity are route-dependent release attributes. Independent batches, predefined acceptance criteria, and stability-indicating studies are needed to quantify batch-to-batch variability and detect aggregation, leakage, charge drift, release changes, or chemical conversion during storage. These attributes should be linked to critical process parameters such as raw-material grade, order of addition, mixing or homogenization energy, solvent removal, purification, drying, and sterilization. This control strategy follows the quality-by-design and risk-management principles described in ICH Q8(R2), ICH Q9(R1), and ICH Q10 and should be adapted to the U.S. Food and Drug Administration’s 2022 guidance for drug products containing nanomaterials.\nRegulatory planning should be anchored to a single clinically intended product rather than a broad platform label. Its quality target product profile should specify route, dosage form, release pattern, and the function of any targeting ligand or active carrier. Product classification and dossier scope would depend on whether a carrier, scaffold, siRNA, or co-delivered agent contributes an additional mode of action, for which early consultation with the relevant authority may be warranted. Before first-in-human evaluation, a candidate would generally require a defined manufacturing process, qualified raw materials, clinical-lot specifications and stability data, route-appropriate microbiological controls, and assays capable of quantifying total and free polydatin. These expectations are consistent with the product and process understanding, risk management, and pharmaceutical quality system principles described in ICH Q8(R2), ICH Q9(R1), and ICH Q10, together with the FDA’s 2022 guidance for drug products containing nanomaterials.\nManufacturing changes during scale-up should be evaluated through analytical comparability; changes in particle-size distribution, loading, release, or degradation may require additional pharmacokinetic, distribution, or toxicology bridging. Complex co-delivery systems also require separate control of siRNA, 4-PBA, gold, extracellular-vesicle, or ligand components and justification of each component’s contribution. The CMC package should mature with trial phase, progressing from qualified methods and provisional acceptance limits to validated assays, process validation, commercial-scale reproducibility, and shelf-life specifications [ 115 ]. For nanomaterial- or liposome-based candidates, the U.S. FDA guidance documents on drug products containing nanomaterials (2022) and liposome drug products (2018) provide additional product-specific considerations. These elements provide the product definition needed to interpret subsequent human evidence.\n\nAgainst this product-development framework, human evidence for polydatin comprises oral combination products, topical preparations, and a registered intravenous study. A randomized multicenter study evaluated a palmitoylethanolamide/polydatin combination containing 20 mg polydatin twice daily for 12 weeks in patients with irritable bowel syndrome [ 116 ]. A placebo-controlled pilot study and a subsequent meta-analysis also evaluated micronized palmitoylethanolamide/trans-polydatin combinations in endometriosis-related pain, although these data do not isolate the contribution of polydatin [ 117 , 118 ]. Another study applied creams containing 0.8% or 1.5% polydatin twice daily for six months to manage cutaneous adverse reactions associated with epidermal growth factor receptor inhibitors [ 119 ]. A phase II study of intravenous polydatin for traumatic, hemorrhagic, or septic shock planned to enroll 240 participants and administer 200 mg over two hours once daily for five days. ClinicalTrials.gov currently lists the study status as unknown; the record was last verified in January 2013 as not yet recruiting, and no results have been posted [ClinicalTrials.gov  NCT01780129 ; accessed 26 August 2026].\nTaken together, these studies establish that polydatin-containing interventions have entered human evaluation through oral, topical, and intravenous routes. Although none assessed a polydatin nanocarrier, comparative pharmacokinetic data are available for oral formulations that increase systemic polydatin exposure [ 67 , 71 ], while local matrices provide sustained release and targeted or responsive platforms are supported mainly by cellular uptake and fluorescence-based distribution [ 23 , 25 , 75 ]. Further development should prioritize clinically feasible routes, controls receiving free polydatin by the same route and at a matched dose, quantitative measurement of polydatin and relevant metabolites at intended sites of action, and experimental designs linking exposure to mechanism-specific pharmacodynamic and safety endpoints. These requirements align with translational nanomedicine frameworks emphasizing reproducible product characterization, relevant preclinical models, and scalable manufacturing [ 120 , 121 ].\nThe reported rodent bioavailability gains should not be interpreted as estimates of human exposure. The three direct oral comparisons used different rat doses (10, 50, and 100 mg/kg), formulations, sampling designs, and analytical reporting, whereas the only human bioanalytical study evaluated conventional tablets and yielded sparse measurements of parent polydatin. Cross-study ratios therefore cannot bridge species or products. A human program should instead compare the intended clinical formulation with a route- and dose-matched free-polydatin reference and quantify parent polydatin, resveratrol, and major glucuronide and sulfate conjugates over a prespecified sampling window.\nBody-surface-area conversion can provide a provisional human-equivalent dose for planning, but it is not a clinical dose recommendation and does not account for formulation-dependent absorption, metabolite formation, or carrier disposition [ 122 ]. This limitation should be considered alongside the body-surface-area approach described in the U.S. FDA’s 2005 guidance on estimating the maximum recommended starting dose in initial adult volunteer studies. Starting-dose selection should integrate route-appropriate repeat-dose toxicology, local tolerance, systemic exposure margins, and the pharmacological contribution of every active carrier or co-delivered component. For oral candidates, exposure matching is more informative than nominal mg/kg conversion; for injectable or local products, carrier burden, infusion or application volume, release rate, and tissue exposure require separate justification.\nEarly PK/PD studies should prespecify analytes and mechanism-linked biomarkers. Plasma polydatin, resveratrol, and principal conjugates should be measured alongside formulation-relevant free and total drug where feasible. Biomarkers should match the indication: lactate clearance, vasopressor requirement, and Sequential Organ Failure Assessment trajectories for shock; serum creatinine, cystatin C, and urinary injury markers for renal disease; pulmonary function, imaging, and fibrosis markers for lung disease; and endothelial function, perfusion, or neurovascular imaging for vascular indications. These endpoints should be analyzed against measured exposure rather than assigned dose alone. This exposure-response approach is consistent with the dose–response principles described in ICH E4.\nWe propose a staged roadmap beginning with a single development candidate supported by batch-reproducible CMC, stability, route-appropriate repeat-dose toxicology, and comparative preclinical PK/PD. Phase I or an early patient study, as appropriate to the indication, would evaluate single and multiple doses, tolerability, complete analyte pharmacokinetics, and formulation-specific risks; food-effect testing would be relevant for oral products and infusion reactions for parenteral products. A Phase Ib/IIa proof-of-mechanism study could then test an indication-specific biomarker and exposure-response hypothesis before larger Phase II efficacy trials. This progression is consistent with the nonclinical-to-clinical expectations described in ICH M3(R2) and the quality-by-design principles for clinical studies in ICH E8(R1). The unupdated status and absence of posted results for  NCT01780129  prevent that registry record from serving as a human exposure or efficacy benchmark.\nUnder this sequence, a nanocarrier should advance only if it offers a measurable advantage over a simpler non-nano formulation at a clinically feasible dose and route. A targeted system would additionally require quantitative drug distribution or target-cell exposure rather than fluorescence alone. These criteria also frame the broader research priorities below.\n\nPolydatin is a chemically defined stilbenoid glucoside whose reported preclinical actions involve redox regulation, mitochondrial and endoplasmic reticulum homeostasis, inflammatory signaling, regulated cell death, autophagy, and mitophagy. Its glucose moiety distinguishes polydatin from resveratrol in physicochemical properties, intestinal handling, and metabolic fate. However, physical-form-dependent dissolution, intestinal hydrolysis, phase-II metabolism, and low oral bioavailability complicate the achievement of predictable exposure [ 14 , 15 , 36 ]. Delivery systems can therefore serve functions beyond solubility enhancement by modifying release, systemic exposure, tissue distribution, cellular uptake, and local residence. Because many reported mechanisms involve intracellular signaling proteins and organelles, total tissue accumulation alone does not establish pharmacologically relevant delivery ( Figure 5 ).\nThe available evidence demonstrates several formulation-specific advantages, although their strength varies across administration routes and disease models. Oral nanoliposomes and fucoidan-based nanoparticles increased systemic polydatin exposure, and the fucoidan systems linked pharmacokinetic data and limited tissue-distribution measurements with functional, histological, and molecular responses in cisplatin-induced kidney injury [ 67 , 71 , 72 ]. Dose-matched studies also reported stronger pharmacodynamic responses than free polydatin in diabetes, intestinal injury, and developing pulmonary fibrosis [ 21 , 69 , 70 ]. Local matrices sustained polydatin release and supported tissue-repair responses, whereas targeted and responsive systems modified cellular uptake or release under defined experimental conditions [ 23 , 25 , 75 ]. For intravenous, local, targeted, and responsive systems, however, quantitative measurement of polydatin in target tissues and relevant cell populations remains uncommon. Pharmacological responses to active carriers and multicomponent formulations must also be interpreted as effects of the complete formulation unless appropriate component controls are included. The current evidence therefore distinguishes successful formulation construction from demonstrated changes in polydatin exposure and from evidence that those changes strengthen mechanism-related pharmacodynamic responses.\nRational carrier and combination design should begin with a defined route-specific barrier rather than with material novelty alone. Carrier composition, release trigger, ligand, and co-delivered agent should each have a prespecified function that can be tested with free-polydatin, blank-carrier, ligand-free, and component-specific controls. Multicomponent systems should be advanced only when the combined design produces an interpretable exposure or pharmacodynamic advantage that cannot be achieved by a simpler formulation. This approach would reduce redundant carrier variants and make formulation selection more closely aligned with the intended clinical indication.\nFurther development should prioritize formulations supported by reproducible evidence of route-appropriate release, altered exposure, or advantages over dose- and route-matched free polydatin. Simultaneous measurement of polydatin, resveratrol, and major conjugated metabolites in plasma and target tissues would clarify how delivery systems alter disposition and metabolic conversion. Studies proposing intracellular delivery should additionally measure cellular polydatin exposure or target engagement and relate these findings to mechanism-specific pharmacodynamic endpoints. Repeated-dose safety, immune responses, material degradation and clearance, batch reproducibility, storage stability, and scalable manufacturing should be evaluated for the formulations selected for further development. No study identified in this review has evaluated a polydatin delivery system in a cardiovascular disease model, despite preclinical evidence for free polydatin in endothelial dysfunction, cardiac remodeling, and atherosclerosis [ 38 , 60 , 96 , 110 ]. Delivery studies in neurovascular disorders are likewise sparse despite a broader preclinical literature on free polydatin [ 6 ]. Extending cardiovascular and neurovascular delivery research should address a defined exposure, cellular-delivery, blood-brain-barrier, or targeting limitation rather than rely on pharmacological activity alone. Machine-learning approaches may eventually assist formulation optimization and prediction of critical quality attributes, but reliable application will require standardized formulation, biodistribution, PK/PD, and safety datasets that are not yet available for polydatin delivery systems [ 123 , 124 , 125 ]. Human studies have evaluated oral combinations and topical preparations containing polydatin [ 116 , 119 ], while the intravenous shock trial remains of unknown status and without posted results [ClinicalTrials.gov  NCT01780129 ; accessed 26 August 2026]; no polydatin nanocarrier has yet undergone clinical evaluation. Candidate selection should therefore favor formulations that combine reproducible manufacturing, route-specific exposure, mechanism-related pharmacodynamic effects, and an acceptable safety profile, with a measurable advantage over free polydatin.\n\nPolydatin delivery research has progressed from solubility-oriented formulation to systems intended to modify exposure, release, distribution, cellular uptake, and local residence. The strongest comparative evidence remains a small oral pharmacokinetic subset; most injectable, targeted, responsive, and local platforms are supported primarily by formulation, imaging, or pharmacodynamic data rather than quantitative drug exposure. Translation will therefore depend on selecting a route-appropriate product, demonstrating reproducible manufacture and safety, measuring polydatin and its metabolites, and showing an exposure-linked advantage over matched free drug. These priorities provide a practical basis for advancing the most credible candidates toward clinical evaluation.","source_license":"CC-BY-4.0","license_restricted":false}