Evaluation of Modified Chitosan Nanomicelles with Carboplatin for Effective Treatment Against Ovarian Cancer Cell Types in Murine Model

Abstract

Introduction

The clinical efficacy of platinum-based chemotherapy in ovarian cancer is limited by insufficient drug retention within the peritoneal cavity and associated systemic toxicity. This study aimed to develop hydrophobically modified chitosan nanomicelles as a carrier system for localized carboplatin delivery.

Methods

Hydrophobically modified chitosan nanomicelles were prepared using an oil-in-watr emulsification technique and characterized in terms of their physicochemical properties, stability, encapsulation efficiency, and drug release behaviour. In vitro studies were performed to assess biocompatibility and cellular internalization in MUC1-expressing SK-OV-3 ovarian cancer cells. The preliminary in vivo antitumour response to the carboplatin-loaded nanomicelles was further evaluated in a subcutaneous murine ovarian cancer xenograft model.

Results

The obtained nanomicelles exhibited a mean hydrodynamic diameter of approximately 200 nm and an encapsulation efficiency of 44%. No quantifiable carboplatin release was detected within 24 h under the tested conditions; however, the absence of a quantifiable signal cannot distinguish limited early release from release below the quantifiable range of the analytical method. The nanomicelles demonstrated preliminary in vitro cytocompatibility in SK-OV-3 cells, with mean cell viability remaining above 70% under the tested conditions, and intracellular localization of the nanomicelles in MUC1-expressing SK-OV-3 cells was observed by confocal microscopy. At the end-point evaluation, tumours were significantly smaller in the nanomicelle-treated group than in the free-carboplatin group (p = 0.03), whereas exploratory longitudinal analysis of tumour-growth AUC showed no significant between-group difference (p = 0.67).

Conclusion

Hydrophobically modified chitosan nanomicelles represent a feasible carrier platform for localized carboplatin delivery and warrant further investigation for potential intraperitoneal application. The observed drug encapsulation, lack of quantifiable carboplatin release during the 24-h study period, preliminary in vitro cytocompatibility in SK-OV-3 cells, cellular internalization, and preliminary antitumour response provide proof-of-concept support for continued evaluation of this carrier system.

Graphical Abstract

Introduction

Ovarian cancer remains one of the most challenging malignancies in modern oncology due to its high incidence, late-stage diagnosis, and aggressive clinical behaviour. Globally, approximately 310,000 women are diagnosed with ovarian cancer each year, with particularly elevated incidence rates observed in Eastern and Central Europe.¹ The asymptomatic nature of early-stage disease and the non-specificity of clinical manifestations—such as abdominal distension, bloating, and unintentional weight loss—often result in diagnosis only after the disease has progressed to an advanced stage. At these later stages, the cancer is frequently characterised by widespread peritoneal dissemination and significant residual disease following cytoreductive surgery, leading to markedly reduced five-year survival rates.^2,3 Despite advances in surgical techniques and chemotherapeutic regimens, recurrence remains a major clinical challenge, underscoring the urgent need for more effective and targeted therapeutic strategies.

Standard treatment for advanced ovarian cancer typically involves optimal cytoreductive surgery followed by adjuvant chemotherapy, most commonly with platinum-based compounds such as carboplatin. Although initial responses are often favourable, with remission rates reaching up to 80%, the majority of patients eventually experience relapse. This is largely attributed to the presence of microscopic metastatic foci within the peritoneal cavity that evade complete surgical removal. In addition, the therapeutic efficacy of carboplatin is limited by its physicochemical properties. Due to its low molecular weight and high-water solubility, free carboplatin rapidly diffuses into the systemic circulation and surrounding tissues, resulting in insufficient local drug retention at the tumour site. This uncontrolled distribution reduces therapeutic efficacy while increasing off-target toxicity in healthy tissues.

To address these challenges, researchers have increasingly focused on developing advanced drug delivery systems to enhance the localisation and controlled release of anticancer drugs directly at the tumour site. Encapsulation of such drugs within micro- or nanocarriers represents a promising strategy in this regard.⁴ Such delivery systems are designed to improve the drug’s pharmacokinetic profile, ensuring sustained and targeted release while minimising systemic exposure and adverse side effects. One of the most promising materials for constructing these carriers is chitosan, a naturally derived polysaccharide obtained through the deacetylation of chitin—a biopolymer found in the exoskeletons of crustaceans and the cell walls of fungi.⁵

Chitosan offers several advantages, making it an ideal candidate for drug delivery applications. It is biodegradable, non-toxic, and biocompatible, which minimises the risk of adverse reactions following administration.^6,7 Perhaps most notably, chitosan possesses exceptional mucoadhesive properties. This mucoadhesiveness arises from its cationic nature, which facilitates strong electrostatic interactions with negatively charged mucosal surfaces.⁸ Such interactions are particularly significant in the context of ovarian cancer, where the tumour microenvironment is often enriched with mucins – a family of high-molecular-weight glycoproteins that play a critical role in both normal cellular functions and oncogenic processes. Despite these favourable properties, unmodified chitosan has limitations as a nanocarrier, particularly related to its physicochemical properties and limited ability to form stable self-assembled nanostructures under physiological conditions.^9,10 Hydrophobic modification of chitosan can overcome some of these limitations by introducing amphiphilic character, promoting self-assembly into nanoscale structures and enabling modulation of their physicochemical and drug-loading properties.^11–13

Mucins, such as MUC1, are typically expressed on the apical surface of epithelial cells, where they serve as a protective barrier against pathogens, mechanical damage, and inflammatory insults.¹⁴ However, in ovarian cancer, the expression pattern of mucins is dramatically altered.¹⁵ Overexpression and aberrant glycosylation of mucins have been observed in many ovarian tumours, resulting in the redistribution of these glycoproteins across the entire cell surface, rather than confining them to the apical region. This phenomenon contributes to the enhanced metastatic potential of the tumour cells by facilitating detachment and dissemination and presents a unique opportunity for targeted drug delivery. The high density of mucins on the surface of ovarian cancer cells can be exploited to achieve controlled and preferential binding of mucoadhesive carriers, such as those constructed from chitosan, thereby increasing the local concentration of encapsulated carboplatin at the tumour site.⁸

The conventional delivery of free carboplatin has significant drawbacks. The rapid systemic absorption of free carboplatin results in a transient tumour exposure to suboptimal drug concentrations, thereby limiting its cytotoxic potential against dispersed cancer cells. Moreover, the non-specific distribution of the free drug increases the likelihood of damage to healthy tissues, contributing to a spectrum of systemic side effects that can severely compromise patient outcomes. This trade-off between effective tumour targeting and minimising collateral toxicity underscores the necessity for innovative drug delivery systems that can reconcile these competing demands.

Chitosan-based nanocarriers have been investigated as potential platforms for localized drug delivery owing to their biocompatibility, biodegradability, and mucoadhesive properties.^6,7 These characteristics provide a rationale for exploring chitosan as a carrier for localized carboplatin delivery. Encapsulation within a polymeric carrier may modify drug distribution and release kinetics compared with the free drug; however, the magnitude and duration of these effects depend on the specific carrier composition and drug–polymer interactions.^16,17 The mucoadhesive properties of chitosan may additionally promote interactions with mucin-rich biological surfaces and potentially contribute to local carrier retention.⁸ In the context of ovarian cancer, these properties provide a rationale for investigating chitosan-based systems for intraperitoneal carboplatin delivery. However, whether these characteristics translate into prolonged carboplatin retention, reduced systemic exposure, or improved therapeutic efficacy requires direct experimental evaluation.

Despite these advantages, the application of hydrophobically modified chitosan nanomicelles for localized carboplatin delivery in ovarian cancer remains largely unexplored. We hypothesize that such nanostructures may enhance local drug retention and therapeutic efficacy through mucoadhesive interactions within the peritoneal tumour microenvironment. To the best of our knowledge, this is the first study investigating hydrophobically modified chitosan nanomicelles as a carrier system for carboplatin delivery. Therefore, the aim of this work was to develop and comprehensively evaluate this system, with particular emphasis on physicochemical properties, in vitro biocompatibility and cellular uptake, as well as preliminary in vivo therapeutic performance.

Materials and Methods

Synthesis of Hydrophobically Modified Chitosan

Chitosan (CH) ChitoClear^® 43000—hqg10 was purchased from Primex ehf Iceland Company (deacetylation degree ~83%, determined by the ¹H NMR, weight average molecular weight: 145.7 kDa determined by GFC). Linoleic acid (LA) (≥99%) 1-ethyl-3-(3- dimethylamino–propyl) carbodiimide hydrochloride (EDC), N-hydroxysuccinimide (NHS), hydrochloric acid conc. (HCl), formic acid conc. (FA), methanol (MeOH), sodium tripolyphosphate (TPP) and coumarin-6 were purchased from Sigma Aldrich Co., Ltd. and used without further purification. Dichloromethane (DCM, analytical grade) was purchased from Chemland, Poland. Carboplatin solution (CPt) 100 mg/mL was delivered from Pomeranian Medical University in Szczecin pharmacy.

Hydrophobically modified chitosan (HMC) was obtained as described in our previous paper.¹⁸ Linoleic acid was first dissolved in 15 mL of methanol (1 equivalent), followed by the addition of EDC (2 equivalents) to initiate activation, which proceeded for 20 minutes. The activated solution was then gradually introduced into a pre-prepared chitosan solution, allowing the reaction to occur over 24 hours. The chitosan solution (0.5 g) was prepared in water, with the pH adjusted to 4.0 using 1M HCl. After the reaction period, the mixture underwent dialysis against water for 48 hours at room temperature using a membrane with a molecular weight cutoff of 3.5 kDa. Subsequently, the product was freeze-dried (Christ Alpha 1–2) for 48 hours, yielding a yellowish, sponge-like solid. Chemical structure of HMC was confirmed by FTIR spectrometry (Alpha, Bruker, 32 scans).

Carboplatin-Loaded Nanomicelle Preparation

Nanomicelles (NMs) were synthesised using an oil-in-water emulsification technique, with process parameters (including pH and solvent ratios) selected based on previous studies on chitosan derivatives and their self-assembly behaviour in mixed solvent systems.¹⁹ HMC was dissolved at a concentration of 1 mg/mL in a mixture of 1% formic or acetic acid and methanol (4:6 v/v). The modified chitosan was placed in a sealed container and dissolved in 1.5 mL of the selected 1% acid at 37°C using a platform shaker at the speed 2 rpm (Heidolph Instruments GmbH and Co. KG, Schwabach, Germany). Next, 6 mL of methanol was added, and stirring continued until the chitosan was fully dissolved. Once complete dissolution was achieved, the remaining acid was added gradually. The resulting solution was filtered through a PES filter (pore size: 0.45 μm), and the pH was adjusted to 5.0 using concentrated NaOH.

The NMs containing CPt were then prepared as follows. Stock carboplatin solution (100 mg/mL) was added to the HMC solution to achieve a final concentration of 1.67 mg/mL. It was transferred to a plastic test tube with a screw cap and shaken for 20 minutes at room temperature (rotation speed: 1000 rpm, IKA Labortechnik, Staufen, Germany). After this, 150 μL of DCM was added, and the mixture was continuously shaken for an additional 30 minutes at 1000 rpm, ensuring the temperature did not exceed 25°C. Afterwards, 1.9 mL of TPP solution (0.5 mg/mL, pH = 5) was gradually added dropwise at 750 rpm. Stirring continued for an additional 30 minutes. The resulting suspension was then divided into centrifuge tubes and centrifuged at 4°C for 30 minutes at 12,000×g (Eppendorf Centrifuge 5415R, Hamburg, Germany). The supernatant was discarded, and the pellet was resuspended in ultrapure water. This was followed by three additional rounds of centrifugation and washing. Finally, the pellet was suspended in water to obtain a nanosuspension of particles.

For coumarin-6-loaded NMs, the dye was incorporated during the emulsification process. Specifically, coumarin-6 was dissolved in the oil phase (DCM) at a concentration of 15 mg/mL. The nanomicelles were then synthesized following the same procedure described above. The presence of the fluorescent dye in the obtained nanomicelles was confirmed by fluorescence microscopy using a Leica DMi8 microscope (Leica Microsystems) equipped with an EL6000 mercury metal halide light source and a Leica DFC7000 GT digital camera. Image acquisition was performed using Leica Application Suite (LAS X) software.

Colorimetric UV-Vis Spectroscopy

Carboplatin encapsulation efficiency was determined using a colorimetric UV-Vis spectrophotometric method based on the formation of a red-colored complex with 1,10-phenanthroline in the presence of ferric chloride, as previously described by Hayat et al.²⁰ Briefly, a calibration curve was prepared using carboplatin standard solutions in the concentration range of 0.14–1.67 mg/mL (n = 6). The upper concentration of 1.67 mg/mL corresponded to the maximum concentration of carboplatin used during the emulsification process. For analysis, the supernatant obtained after the emulsification process was subjected to the same colorimetric reaction and the absorbance of the resulting complex was measured at 510 nm using a NanoQuant Infinite M200 PRO TECAN spectrometric plate reader. The colorimetric reaction was performed using 0.5% ferric chloride and 0.4% 1,10-phenanthroline, followed by incubation at 60°C for 45 min. After cooling to room temperature, 15 mM phosphoric acid was added, and the absorbance was measured against a reagent blank prepared under identical conditions. The relationship between absorbance and carboplatin concentration was described by the following linear equation:

(1)

where y is absorbance and x is concentration (mg/mL). Calibration curve is presented in Supplementary Data file, Figure S1.

The average absorbance for supernatant was 1.0598 (n=6). Therefore, CPt concentration in the supernatant = 0.94 mg/mL and finally, CPt concentration encapsulated in micelles:

(2)

Encapsulation efficiency was calculated based on equation 2:

(3)

Drug loading (DL) was not determined because the CPt-HMC-NMs were characterized as an aqueous nanosuspension and were not isolated or dried to determine the total mass of the drug-loaded nanomicelles required for DL calculation.

Drug Release Analysis

Pellets collected in Eppendorf tubes were resuspended in 200 μL of ultrapure water. The resulting suspensions were combined and transferred into a dialysis membrane (MWCO 1 kDa). The membrane was immersed in 250 mL of ultrapure water and maintained at 37°C under continuous stirring at 200 rpm to provide controlled temperature and mixing conditions. At predetermined time points (1, 2, 3, 4, 5, 6, and 24 h), 1 mL aliquots were withdrawn from the external medium and immediately replaced with an equal volume of fresh ultrapure water to maintain a constant release-medium volume. The collected samples were analysed using a NanoQuant Infinite M200 PRO TECAN spectrometric plate reader at 510 nm after completion of the colorimetric reaction described above. Carboplatin concentration was determined based on a previously established calibration curve described by the linear equation . The cumulative amount of released carboplatin was calculated by accounting for the amount of drug removed with the previously withdrawn samples and was expressed as a percentage of the initial amount of carboplatin in the formulation. As all measured carboplatin concentrations remained below the quantifiable range of the analytical procedure, no quantifiable cumulative drug release was detected during the 24-h study period. All experiments were performed in triplicate (n = 3).

Dynamic Light Scattering (DLS)

The nanomicelles hydrodynamic diameter was analysed using a dynamic light scattering (DLS) instrument (Zetasizer NanoZS, Malvern Instruments Ltd., Malvern, UK). Measurements were conducted at 37°C with a backscatter detection system set at a 173° angle, with each sample analysed six times. The excitation source used was a helium-neon laser with vertical polarisation, operating at a wavelength of 633 nm. The hydrodynamic diameter was monitored over time, with assessments taken at 24 hours in water and peritoneal fluid 37°C. Between measurements, the samples were incubated at 37°C using a platform shaker (Heidolph Instruments GmbH and Co. KG, Schwabach, Germany) to maintain consistent conditions.

Cell Culture (SK-OV-3)

Human ovarian carcinoma SK-OV-3 cells (ECACC 91091004) were obtained from the European Collection of Authenticated Cell Cultures (ECACC). Cells were cultured in McCoy’s 5A Modified Medium (Corning) supplemented with 10% heat-inactivated fetal bovine serum (FBS; EURx), 2 mM L-glutamine (Sigma-Aldrich), and 100 IU/mL penicillin and 100 µg/mL streptomycin (Sigma-Aldrich). Cells were maintained at 37 °C in a humidified atmosphere containing 5% CO₂ and were routinely screened for mycoplasma contamination using the MycoAlert^® PLUS Mycoplasma Detection Kit (Lonza, Basel, Switzerland). Experiments were performed using cells at passage 3. Upon reaching confluency, cells were trypsinized with 1% trypsin solution (Sigma-Aldrich) for 5 min at 37°C and centrifuged at 120 × g for 10 min at 4°C. The cell pellet was resuspended, counted, and centrifuged again. A total of 40 × 10⁶ cells were suspended in 1 mL of ice-cold Matrigel^® Matrix (ECM Gel from Engelbreth–Holm–Swarm murine sarcoma; Sigma-Aldrich). This procedure was prepared in multiple independent preparations under identical conditions to obtain sufficient material for all experimental animals (n = 34). From the moment of Matrigel addition, the cell suspension was kept on ice. Aliquots of 100 µL containing 4 × 10⁶ cells were loaded into pre-cooled insulin syringes and administered subcutaneously to BALB/c Nude mice for xenograft establishment.

Cell Viability Assay (WST-1)

SK-OV-3 cells were cultured as above, and the biocompatibility (cell viability) of the nanomicelles was evaluated using the Cell Proliferation Reagent WST-1 assay (Roche Diagnostics, Mannheim, Germany). The WST-1 test is based on reduction of the tetrazolium salt to formazan, which final amount is directly correlated to the number of metabolically active cells. In the present study, SK-OV-3 (2 x 10³) cells were seeded in 96-well plate and then cultured in standard conditions. After 24 h the culture medium was removed and replaced with fresh medium containing the nanomicelles at the final concentration range equal to 1–0.002 mg/mL for 24, 48 or 72 h. The cells without nanomicelle (only culture medium exposition) were used as control, and tested nanomicelle in medium (without cells) as a blank. After each respective exposure period (24, 48, or 72 h), WST-1 reagent was added, incubated with the cells for 30 min, and absorbance was measured at 450 nm (with 620 nm background correction) using a spectrophotometric microplate reader (Infinite 200 Pro, Tecan, Männedorf, Switzerland). Cell viability was calculated using the following formula: [(A_test – A_blank)/(A_control − A_blank)] × 100%. The WST-1 assay was performed in three independent experimental runs for the 24- and 48-h exposure periods and in two independent experimental runs for the 72-h exposure period. Each independent experiment was conducted in technical triplicate. Technical replicates were averaged within each independent experiment before statistical analysis.

Cellular Uptake (Immunofluorescence Staining)

Fixed cell culture slides were rinsed with PBS (Sigma-Aldrich) and the cells were permeabilized for 20 minutes in 0.5% TRITON X100 solution (Sigma-Aldrich). The detergent was then rinsed with PBS, and the slides were incubated with blocking serum (2.5% horse serum in PBS (Thermo-Fischer) for 20 minutes at room temperature. The slides were then incubated with appropriately diluted monoclonal mouse anti-human MUC1antibody (ThermoFisher Scientific, Waltham, MA, USA) overnight at 4°C in a humid chamber. The primary antibody was then washed off with PBS. The cells were then incubated with a secondary antibody coupled to a fluorochrome: anti mouse produced in rabbit, Texas Red (Sigma, Merck Millipore, Darmstadt, Germany) for 1 hour at room temperature in a humid chamber in the dark. The slides were rinsed again in PBS and incubated with DAPI (Merck Millipore, Darmstadt, Germany) for 20 minutes at room temperature to visualize cell nuclei. After a final wash in PBS, the prepared slides were mounted in fluorescence medium (Dako). The slides were examined in FV1000 confocal microscopy system (Olympus, Germany) coupled to an IX81 inverted microscope (Olympus, Germany). Images were acquired using multichannel imaging: a 488 nm laser for FITC, 595nm for Texas Red and a 405 nm diode laser for DAPI.

The in-vivo Studies Using Animal Model

Study was performed on 34 female nude mice (Crl:NU-Foxn1nu strain) 6–8 weeks old purchased from Charles River Laboratories (Germany). Mice were housed under controlled conditions of optimum temperature (21°C), ventilation, and light timed to follow a circadian rhythm (12:12 h). The mice were maintained in sterile cages within an IVC (Individually Ventilated Cage) rack system (providing 54 air changes per hour), equipped with transparent, high-quality polysulfone cages with a floor area of 550 cm² and a height of 14 cm, each fitted with an external filter and appropriately labeled. Five mice were kept per cage, with environmental enrichment provided on dust-free aspen bedding, which was replaced at least once per week (or as required). In cases of aggressive behavior, animals were separated into individual cages; the cage design allowed for visual contact between separated conspecifics. All animals had free access to tap water and commercial diet for the maintenance of adult rodents—Labofeed H (containing net energy 12.8 MJ/kg in proportion: 60% carbohydrates, 30% proteins, 10% fat) (Morawski, Poland)) ad libitum. After 1 week acclimatization in animal facility mice were injected subcutaneously, with SK-OV-3 cells (1x10⁶ cells/mice, 25 μL) in Matrigel (Corning Matrigel Basal membrane matrix, Merck, Burlington, MA, USA) for tumour development induction.²¹ Mice were injected subcutaneously with SK-OV-3 cells for tumour development induction. Although the nanomicelles system was designed for intraperitoneal administration, a subcutaneous tumour model was employed to enable controlled and reproducible assessment of tumour growth and treatment response. Tumour growth was assessed at the injection site by performing biweekly caliper measurements, conducted concurrently with routine body-weight monitoring. To estimate tumour size, the longest longitudinal diameter (length) and the largest transverse diameter (width) were measured in conscious mice using external calipers. For longitudinal presentation, time was expressed as days from the first tumour measurement (day 0). This time point preceded treatment initiation in animals in which the tumour had not yet reached the minimum size required for treatment. Therefore, the longitudinal observation period was longer than the six-week treatment period. Group data were summarized as mean ± SEM over the common observation period with complete measurements in both treatment groups (Figure S2).

Body weight was recorded longitudinally throughout the observation period. For each animal, body-weight change was expressed as the percentage change relative to the first body-weight measurement included in the longitudinal analysis (day 0), which was used as the baseline for longitudinal follow-up. Group-level changes were summarized descriptively and compared between the treatment groups.

The study was conducted in accordance with the National Institutes of Health (NIH) Guidelines for the Care and Use of Laboratory Animals and the European Community Council Directive on the Care and Use of Laboratory Animals. The study was approved by the Local Ethics Committee for Experiments of Animals in Poznań (No. 59/2023, dated 13 October 2023). The number of animals that participated in the study was adopted according to the requirements of DIRECTIVE 2010/63/EU of the European Parliament and of the Council of 22 September 2010 on the protection of animals used for scientific purposes (3R rule: Replacement, Reduction, Refinement).

Tumour Treatment

After tumour establishment was confirmed, animals were randomly assigned to carboplatin and carboplatin-loaded nanomicelles treatment groups. Animal well-being was monitored throughout the study, including assessment of food intake and potential signs of chemotherapy-related toxicity; however, no adverse effects were observed, and administration of antiemetic agents was not required.

Carboplatin and carboplatin-loaded nanomicelles were administered subcutaneously in the vicinity of the tumour (minimum size 200 mm²) at a dose of 50 mg/kg body weight once weekly for six consecutive weeks, as previously described by Brodeur et al.²¹ This dosing regimen was selected based on prior studies demonstrating therapeutic efficacy over shorter treatment cycles and was extended to six cycles to better reflect clinically relevant chemotherapy schedules.

Animals were monitored for body weight changes, food intake, and potential treatment-related side effects throughout the study. Longitudinal body-weight analysis included animals for which complete body-weight data were available over the common observation period (NC, n = 15; C, n = 13); two animals were excluded from this analysis because body-weight measurements were incomplete or unavailable. At the end of the treatment period, animals were euthanized using Mobital (Biowet, Puławy, Poland). Tumours were excised, weighed, and processed for further histological analysis.

Histopathology and Immunohistochemistry

Ki-67 Immunohistochemistry

Sections of tumour (4 μm thick) were hydrated, and heat epitope retrieval (HIER) was performed in microwave oven in citrate buffer pH=6 (Dako Retrieval Solution, Dako, Agilent, Santa Clara, CA, USA). After cooling to room temperature (RT), the activity of endogenous peroxidase was blocked with BLOXALL Endogenous Enzyme Blocking Solution (ImmPRESS HRP Universal (Horse Anti-Mouse/Rabbit IgG) PLUS Polymer Kit, Peroxidase, Vector Laboratories, USA) washed twice with PBS and further incubated with 2.5% Normal Horse Serum (ImmPRESS HRP Universal PLUS Polymer Kit, Peroxidase, Vector Laboratories, USA). Further, slides were incubated with primary monoclonal antibodies: mouse anti- Ki-67 M7240 MIB-1 (Dako, Agilent, Santa Clara, CA, USA) 1h in RT. After double wash in PBS slides were incubated with ImmPRESS Universal Antibody Polymer Reagent (ImmPRESS HRP Universal PLUS Polymer Kit, Peroxidase, Vector Laboratories, USA), after washing in PBS reaction was visualised with ImmPACT DAB EqV Substrate (ImmPRESS HRP Universal PLUS Polymer Kit, Peroxidase, Vector Laboratories, USA). Slides were counterstained with hematoxylin (Harris modified Hematoxylin, Sigma, Merck, Darmstadt, Germany) and mounted in PathoMount (MAR-FOUR, Poland) mounting medium and evaluated under an Olympus IX81 inverted microscope (Olympus, Germany). Micrographs were collected with CellSens software (Olympus, Germany). Positive reaction was described as yellow to brown pigmentation found in nuclei. All stained nuclei were considered positive, irrespective of intensity.²² Every sample was evaluated by two researchers independently.

Histopathological Analysis

For histopathological tumour evaluation sections were stained with H-E method according to standard protocol. Briefly, after rehydration in alcohol (99.9% to 70%) slides were placed in distilled water. Harris modified hematoxylin (Sigma Aldrich, Merck KGaA, Darmstadt, Germany) was used for acidic compounds staining for 1 minute RT, after washing in running tap water and distilled water slides were immersed in eosin aqueous solution (Sigma Aldrich, Merck KGaA, Darmstadt, Germany) for basic compounds staining for 1 minute RT. After dehydration in alcohols and clearance in Xylene samples were mounted in PathoMount (MAR-FOUR, Poland) mounting medium and evaluated under an Olympus IX81 inverted microscope (Olympus, Germany). Micrographs were collected with CellSens software (Olympus, Germany). Each sample was evaluated independently by two researchers.

For semiquantitative evaluation of Ki-67 immunoreactivity, a modified immunoreactive score (IRS) was used as previously described by Al-Khan et al.²³ Each tumour sample was evaluated in five randomly selected fields of view using a ×40 objective. The IRS for each evaluated field was calculated by multiplying the staining intensity score (0–3) by the score corresponding to the estimated proportion of Ki-67-positive cells (0–4). The staining intensity score was defined as follows: 0, negative; 1, weak; 2, moderate; and 3, strong staining. The proportion score was defined as follows: 0, no positively stained cells; 1, 1–10%; 2, 11–50%; 3, 51–80%; and 4, 81–100% positively stained cells. The resulting IRS ranged from 0 to 12, with a score of 0 classified as negative, 1–6 as weakly positive, and 7–12 as strongly positive.

Statistical Analysis

Data are presented as mean ± SD unless otherwise specified. Longitudinal tumour-growth data are presented as mean ± SEM. Statistical analyses were performed using GraphPad Prism 9.0 (GraphPad Software, USA). Data normality was assessed using the Shapiro–Wilk test. For normally distributed data, comparisons between groups were performed using the Student’s t-test, while the Mann–Whitney U-test was applied for non-normally distributed data. All tests were two-tailed. A p-value < 0.05 was considered statistically significant. As an exploratory sensitivity analysis of longitudinal tumour burden, the area under the tumour-size–time curve (AUC) was calculated for each animal using the trapezoidal method over the common observation period. AUC values were compared between treatment groups using Welch’s t-test. Body-weight change was calculated for each animal as the percentage change from the first body-weight measurement included in the longitudinal analysis (day 0) to the end of the common observation period and was compared between treatment groups using Welch’s t-test. The number of independent experiments (n) is indicated in the figure captions.

Ki-67 IRS values obtained from individual microscopic fields were summarized descriptively and were not treated as independent biological replicates for between-group statistical testing.

For the WST-1 assay, technical triplicates were averaged within each independent experimental run before statistical analysis. Differences among nanomicelle concentrations within the 24- and 48-h exposure periods were assessed using the Friedman test because measurements at different concentrations were obtained within the same independent experimental runs. Because only two independent experiments were available for the 72-h exposure period, concentration-dependent differences at this time point were evaluated descriptively. Differences among exposure times were explored separately at each concentration using the Kruskal–Wallis test. Statistical significance was defined as p < 0.05. Given the limited number of independent experiments, particularly at 72 h, these analyses were interpreted cautiously.

Results

Nanomicelles Characterization

The chemical structure of HMC was confirmed before nanomicelles preparation. Figure 1 presents spectra of chitosan and HMC with characteristics regions of selected bands. The bands marked as (A) correspond to the presence of aliphatic groups. This group includes the methylene group (-CH₂-) and the ethyl group (—CH₂CH₃) whose vibrations appear at a wavenumber region 2926–2850 cm⁻¹. The increase in absorbance observed for HMC is related to the grafted fatty acid molecules. Another characteristic band (B) at a wavenumber of approximately 1740 cm⁻¹ corresponds to the vibrations of the C=O double bond in the ester group (—COOR’), formed after the acid residue attaches to oxygen (the O-acylated part of the derivative). The (C) band at approximately 1630 cm⁻¹ also corresponds to vibrations of the C=O bond, originating, in this case, from the amide group (—CONHR’), which appeared after the attachment of the aliphatic residue from the acid to the nitrogen atom (the N-acylated part of the derivative), confirming successful linoleic acid grafting. The increase in the share of alkyl groups originating from aliphatic chains of fatty acids in the derivative macromolecules is also reflected in the absorption band at 1435 cm⁻¹, which is shifted and sharpened for subsequent derivatives in the series. A comprehensive analysis of the IR spectra has been previously reported by Piegat et al,¹⁸ focusing on the chemical structure of the materials obtained under different reaction conditions.

Figure 1.

FTIR spectra of neat chitosan (blue) and hydrophobically modified chitosan (HMC; green) with characteristic absorption bands assigned to aliphatic groups (A), ester groups (B), and amide groups (C), confirming successful modification.

As shown in Figure 2, upon carboplatin loading (green bars), a notable shift in the particle size distribution is observed. The mean hydrodynamic diameter of the CPt-HMC-NMs was 191.7 ± 11.7 nm, compared with 220.8 ± 7.2 nm for unloaded HMC-NMs (n = 6). Thus, the incorporation of carboplatin was associated with an approximately 13.2% reduction in mean hydrodynamic diameter, and the difference was statistically significant (p = 0.0021). The corresponding PDI values were 0.331 ± 0.021 and 0.340 ± 0.035 for CPt-HMC-NMs and HMC-NMs, respectively. The CPt-HMC-NMs exhibit a slight shift toward smaller sizes, accompanied by an increase in intensity in the 100–250 nm range. The observed shift towards smaller hydrodynamic sizes indicates a change in the size distribution following carboplatin incorporation. However, DLS measurements alone do not allow the molecular basis of this change to be determined. Possible changes in polymer conformation, hydration, or intermolecular interactions may contribute to the observed difference, but these possibilities were not directly investigated in the present study. In high-concentration CPt solution, hydrogen bonds exist between the ammonia molecules of one complex and the oxygen atoms of the dicarboxylate carboplatin ligand of adjacent complexes.²⁴ In HMC and CPt mixtures, those interactions might form between the ammonia and oxygen parts of CPt and the free -NH₃ and -OH groups of chitosan macromolecules. Furthermore, CPt-HMC-NMs exhibit higher intensity peaks in the lower nanometer range (~120–250 nm), implying probably a more homogenous distribution with potentially improved colloidal stability.

Figure 2.

Size distribution of unloaded nanomicelles (grey) and carboplatin-loaded nanomicelles (CPt-HMC-NMs, green); measurements performed by DLS (n = 6).

In peritoneal fluid, the DLS intensity-weighted size distribution shifted towards larger apparent hydrodynamic sizes, with signals extending to approximately 825 nm (Figure 3A and B). This increase in nanomicelle size may result from the surface adsorption of biomolecules (eg, proteins, phospholipids) or osmotic swelling caused by changes in the hydration shell of the polymeric matrix. The presence of endogenous particles in the peritoneal fluid (diameter range: 3–78 nm) further complicates the analysis, as these particles were also detected in the NMs-peritoneal fluid mixture. The interaction between micelles and naturally occurring components in the peritoneal fluid may suggest a potential alteration in the micellar surface chemistry, which may influence aggregation, charge distribution, and drug release kinetics.

Figure 3.

Size distribution profiles obtained by DLS: (A) native peritoneal fluid (PF; grey) and PF containing CPt-loaded nanomicelles (green); (B) comparison of nanomicelles in aqueous suspension (grey) and after incubation in PF (green) (n = 6).

The thermal stability of micelles was evaluated under physiological conditions (37°C). In aqueous suspension, the mean hydrodynamic diameter increased from 212 ± 6 nm at 25°C to 426 ± 56 nm at 37°C, indicating a temperature-driven expansion of the NMs structure. Additionally, samples were incubated at 37°C for 24 h in peritoneal fluid and reanalyzed (Figure 4). Nanomicelles in peritoneal fluid displayed three distinct size populations (range: 2 −9 nm, 31–37 nm, and 200–285 nm), reflecting a more complex transformation process. A population of small-scattering species below 10 nm was detected in the DLS measurements. However, a similar size distribution was observed in the peritoneal fluid control, indicating that these signals cannot be unequivocally attributed to nanomicelle degradation. Although partial nanomicelle degradation could represent one possible explanation, these species may also correspond to components inherent to the peritoneal fluid or background scattering. Therefore, their origin cannot be determined from the present DLS measurements alone, and complementary analytical techniques would be required to confirm nanomicelle degradation.

Figure 4.

Size distribution of CPt-loaded nanomicelles before (green) and after (pink) 24 h incubation in peritoneal fluid (n = 6).

To evaluate the storage stability of the developed system, the hydrodynamic diameter of both unloaded (HMC-NMs) and CPt-loaded nanomicelles (CPt-HMC-NMs) was monitored over 40 days in aqueous conditions at 25°C, mimicking typical storage conditions (Figure 5). The unloaded micelles exhibited a gradual increase in size over time, which can be attributed to progressive hydration and swelling of the polymer matrix, ultimately leading to reduced structural stability. In contrast, CPt-loaded nanomicelles demonstrated a markedly slower increase in hydrodynamic diameter and maintained a more stable size distribution throughout the incubation period. The observed change in hydrodynamic diameter may be consistent with changes in the chitosan-based shell associated with carboplatin incorporation; however, the molecular nature of drug–polymer interactions was not directly investigated in the present study. CPt-loaded nanomicelles exhibited greater stability of hydrodynamic diameter during the investigated storage period compared with unloaded HMC-NMs, suggesting that carboplatin incorporation may influence the colloidal behavior of the nanomicelles.

Figure 5.

Time-dependent changes in hydrodynamic diameter of unloaded (HMC-NMs; black) and carboplatin-loaded (CPt-HMC-NMs; green) nanomicelles during storage at 25°C. Data are presented as mean ± SD (n = 6).

The encapsulation efficiency of the developed system was determined to be 44%. After 24 h, carboplatin in the external medium remained below the quantifiable range of the analytical procedure under the conditions used. This finding indicates that carboplatin release was not quantifiable within the first 24 h; however, the available data do not allow discrimination between delayed release, strong drug retention within the polymer matrix, and limitations of the analytical method. Therefore, extended-release studies using more sensitive and appropriately validated analytical techniques are required to characterize the release kinetics.

In vitro Studies

Cell Viability

To assess the cellular response to the carboplatin-loaded nanomicelles, SK-OV-3 cells were exposed to the nanostructures at final concentrations ranging from 1.0 to 0.002 mg/mL for up to 72 h. Mean cell viability remained above 70% at all tested concentrations and exposure times (Figure 6). The lowest mean cell viability was observed after 72 h at the highest tested nanomicelle concentration of 1 mg/mL (74.0 ± 2.8%). No significant concentration-dependent differences were detected after 24 h (Friedman test, χ²(9) = 10.91, p = 0.282). An overall effect of concentration was detected after 48 h (Friedman test, χ²(9) = 19.65, p = 0.020); however, post-hoc multiple comparisons did not identify statistically significant differences between individual concentration pairs after correction for multiple testing. Because only two independent experiments were available for the 72-h exposure period, these data were evaluated descriptively. Exploratory comparisons among the 24-, 48-, and 72-h exposure periods showed no statistically significant differences at any individual concentration (all p > 0.05). According to EN ISO 10993–5, maintenance of cell viability at or above 70% is generally considered indicative of acceptable cytocompatibility. However, as the assay was performed exclusively using SK-OV-3 ovarian cancer cells, these findings provide only a preliminary indication of in vitro cytocompatibility and should not be interpreted as evidence of general biocompatibility. As cell viability did not decrease to 50% within the investigated concentration range, an IC50 value could not be determined under the tested conditions.

Figure 6.

Cell viability of SK-OV-3 cells evaluated by WST-1 assay after 24, 48, and 72 h exposure at concentrations ranging from 1.0 to 0.002 mg/mL.

Cellular Uptake

Confocal microscopy demonstrated intracellular localization of coumarin-6-labelled nanomicelles in SK-OV-3 cells (Figure 7). Co-localization with MUC1-expressing cells was observed; however, the present qualitative experiment does not allow quantification of cellular uptake or establish MUC1-specific targeting.

Figure 7.

Confocal microscopy image of SK-OV-3 cells showing cellular uptake of coumarin-6-labelled nanomicelles (green) and MUC1 expression (red); nuclei stained with DAPI (blue). Scale bar 30 µm.

In vivo Therapeutic Evaluation

Tumour Growth

After the inoculation of SK-OV-3 cells in Matrigel to 34 female nude mice, 4 mice failed to develop the tumour. In the remaining animals the subcutaneous tumours developed, and those with the size of 200 mm2 were used for the study. The tumour size parameters at the end of the experiment are presented in Figure 8. Tumours in mice treated with NC were significantly smaller than those in mice treated with free carboplatin (p = 0.03). As an exploratory sensitivity analysis, cumulative tumour burden over time was additionally assessed by calculating the area under the tumour-size–time curve for each animal over the common observation period. Tumour-growth AUC did not differ significantly between the NC and C groups (p = 0.67). Longitudinal tumour-growth data are presented in Figure S2.

Figure 8.

Tumour size at the end of the experiment in mice treated with free carboplatin (C) and carboplatin-loaded nanomicelles (NC). Data are presented as mean ± SD. Tumours in the NC-treated group were significantly smaller than those in the C-treated group (p = 0.03).

Body-Weight Changes

Body weight remained generally stable during the treatment period in both groups. The mean percentage change from baseline to the end of the common observation period was −1.20 ± 6.02% in the NC group and +0.98 ± 5.09% in the free-carboplatin group. The difference between groups was not statistically significant (p = 0.31). Longitudinal body-weight changes are presented in Figure S3.

Histopathology Analysis

In both treatment groups, the tumours exhibited histopathological features of clear cell carcinoma, with a high proportion of clear cells (Figure 9). Hyperchromatic nuclei were observed in both groups, while nuclear pleomorphism and atypia were limited. Mitotic figures were infrequent in tumour sections from both treatment groups. In the carboplatin-treated group, up to four mitotic figures were observed at ×200 magnification and up to two at ×400 magnification. In the carboplatin-loaded nanomicelle group, no more than two mitotic figures were observed at ×200 magnification and no more than one at ×400 magnification (Figure 9A, B, D and E). Representative images of tumours from the free-carboplatin group are shown in Figure 9A–C, whereas Figure 9D–F shows tumours from the carboplatin-loaded nanomicelle group. Apoptotic-like pyknotic nuclei were observed in tumours from the free-carboplatin group (Figure 9B, insert). Hemosiderin deposits were observed in the tumour of one animal from the carboplatin-loaded nanomicelle group (Figure 9E, insert).

Figure 9.

Histopathological analysis and Ki-67 expression in tumours treated with free carboplatin (C) and CPt-loaded nanomicelles (NC). (A and B) H&E staining of C group; (C) Ki-67 in C group; (D and E) H&E staining of NC group; (F) Ki-67 in NC group. Arrows indicate mitotic figures. Inserts show pyknotic nuclei (B) and hemosiderin deposits (E). Scale bars: 50 µm (A, C, D, F) and 10 µm (B and E).

Ki-67 Expression

Ki-67 immunoreactivity was evaluated semiquantitatively using the immunoreactive score (IRS). Across the evaluated microscopic fields, the mean IRS was 3.76 in the free-carboplatin group and 3.89 in the carboplatin-loaded nanomicelle group, with a median IRS of 4 in both groups. IRS values ranged from 1 to 8 and from 0 to 8, respectively. These values represent descriptive field-level observations and were not treated as independent biological replicates for between-group statistical testing. Representative Ki-67 immunohistochemical staining is shown in Figure 9C and F.

Discussion

This study demonstrates the feasibility of using hydrophobically modified chitosan nanomicelles as a carrier for carboplatin. The formulation showed a mean hydrodynamic diameter of approximately 192 nm after carboplatin incorporation, an encapsulation efficiency of 44%, preliminary in vitro cytocompatibility in SK-OV-3 cells, and microscopically observed intracellular localization. In the preliminary in vivo study, carboplatin-loaded nanomicelles were associated with a reduction in tumour size compared with free carboplatin. These findings support the feasibility of the formulation as a nanocarrier and provide a basis for further investigation of its suitability for localized drug delivery. However, the present data do not establish prolonged drug retention, sustained release, or prolonged intraperitoneal residence, as these parameters were not directly measured.

The physicochemical behaviour of the nanomicelles provides some insight into how carboplatin incorporation and the biological environment may affect the formulation. Carboplatin-loaded nanomicelles exhibited a smaller mean hydrodynamic diameter than unloaded HMC-NMs, together with a greater contribution of particles in the 100–300 nm range. This change may reflect rearrangement or compaction of the polymer structure during drug incorporation. The incorporation of carboplatin may alter the intermolecular interactions and hydration state within the nanomicellar structure, although the molecular basis of these changes was not directly investigated in the present study. The slower increase in hydrodynamic diameter of carboplatin-loaded nanomicelles during storage may similarly indicate changes in polymer hydration or chain mobility following drug incorporation. Previous studies have shown that hydrophobically modified chitosan derivatives can self-assemble into micellar structures and that their size distribution and physicochemical properties are influenced by polymer structure and emulsification conditions.^19,25 Similar changes in physicochemical characteristics have also been reported following carboplatin incorporation into chitosan-based nanoparticles.²⁶

The response of the nanomicelles to peritoneal fluid further demonstrates the importance of the biological environment when assessing their physicochemical behaviour. Incubation in peritoneal fluid resulted in a shift towards larger particle populations, which may be associated with adsorption of endogenous biomolecules, changes in hydration, or aggregation. Such changes are consistent with the known effects of biological media on nanoparticle size and dispersion behaviour.^27,28 The detection of small-scattering species below 10 nm after incubation for 24 h should be interpreted with particular caution. Similar populations were present in the peritoneal fluid control, and therefore the present DLS measurements do not allow these signals to be assigned to nanomicelle degradation. This interpretation is consistent with the recognized limitations of light-scattering measurements in complex biological matrices, where signals from nanoparticles may overlap with those generated by endogenous particulate components.²⁷ Complementary methods would be required to distinguish degradation products from endogenous components of the peritoneal fluid.

The potential interaction of HMC-NMs with mucin-rich environments provides a biological rationale for further investigation of this formulation in ovarian cancer. Chitosan is a cationic polysaccharide with established mucoadhesive properties, and electrostatic interactions between protonated amino groups and negatively charged mucins have been described.⁸ In ovarian cancer, altered expression and glycosylation of mucins, including MUC1, may contribute to the interaction of tumour cells with their surrounding microenvironment.^14,15 Nevertheless, mucoadhesion of the present HMC-NMs was not directly measured, and the current study therefore cannot demonstrate increased residence at the tumour or peritoneal surface. Accordingly, the possible contribution of mucoadhesive interactions to local drug localization should be considered a hypothesis requiring direct evaluation in future studies.^8,29

Carboplatin encapsulation efficiency was 44%, indicating that a substantial fraction of the drug was incorporated into the nanomicellar formulation, although the efficiency remains open to further optimization. The observed value can be considered in the context of other carboplatin-loaded polymeric systems. Previous studies have reported lower encapsulation efficiencies for PLGA nanoparticles and values of a similar order for poly(ε-caprolactone) nanoparticles, whereas higher efficiencies have also been reported for chitosan nanoparticles prepared by ionic crosslinking.^16,17,30 Differences between these systems may reflect the physicochemical properties of carboplatin, polymer architecture, drug–polymer interactions, and the encapsulation method. The moderate encapsulation efficiency obtained here therefore suggests that both formulation composition and processing conditions should be systematically optimized rather than attributing the result to polymer chemistry alone.

The release experiment provides an important, but limited, observation regarding the behaviour of encapsulated carboplatin. Carboplatin concentrations in the external medium remained below the quantifiable range of the analytical procedure throughout the 24-h experiment. The lower concentration of the calibration range was 0.14 mg/mL, whereas formal LOD and LOQ values were not determined. Therefore, the present result should be described as no quantifiable carboplatin release under the tested conditions rather than as evidence of complete drug retention or sustained release. The absence of quantifiable drug in the external medium may be consistent with limited early release, but it may also reflect the sensitivity and validation characteristics of the analytical method. The present experiment cannot distinguish between delayed release, retention of carboplatin within the nanomicellar matrix, and release below the quantifiable range. Consequently, the 24-h experiment should be regarded as an initial assessment of early carboplatin release rather than a complete characterization of release kinetics. Extended studies using a more sensitive and appropriately validated analytical method, together with longer observation periods, are required to determine whether the formulation exhibits prolonged release or retention.

The cellular uptake experiments provide complementary evidence that the nanomicelles can interact with and enter SK-OV-3 cells. Confocal microscopy demonstrated intracellular localization of coumarin-6-labelled nanomicelles, while co-localization with MUC1-expressing cells was observed. The experiment was qualitative and did not include quantitative fluorescence analysis or a free-dye comparator; therefore, it cannot determine the magnitude of uptake or whether nanomicelle encapsulation enhances cellular internalization relative to the non-encapsulated fluorescent marker. Chitosan-based nanoparticles are known to undergo cellular internalization through several endocytic pathways, including clathrin-mediated and caveolae-mediated endocytosis and macropinocytosis.^31,32 The mucoadhesive properties of chitosan may facilitate interactions with mucin-rich cell surfaces without requiring an additional targeting ligand, as chitosan can interact with mucins through a combination of electrostatic and other non-covalent interactions.⁸ However, the observed co-localization with MUC1-expressing cells does not establish a specific MUC1-mediated mechanism of cellular uptake. Further studies using receptor-blocking, competition, or other mechanistic approaches would be required to determine whether MUC1 directly contributes to nanomicelle internalization.

The preliminary in vivo findings provide functional evidence for the carboplatin-loaded formulation. Mice receiving carboplatin-loaded nanomicelles had significantly smaller tumours than animals treated with free carboplatin (p = 0.03), providing preliminary evidence of a difference in tumour response between the treatment groups. However, an exploratory analysis of cumulative tumour burden over the common observation period did not demonstrate a significant between-group difference, indicating that the observed end-point difference should be interpreted cautiously. Body weight remained generally stable in both treatment groups, with no significant difference in percentage body-weight change between mice receiving carboplatin-loaded nanomicelles and those receiving free carboplatin. This finding does not indicate a detectable difference in overall treatment tolerability based on body weight alone.

Descriptive field-level Ki-67 IRS values were similar in magnitude in the two treatment groups; however, individual microscopic fields were not considered independent biological replicates, and no between-group statistical inference was made on the basis of these field-level data. Therefore, the observed reduction in tumour size should not be interpreted as being accompanied by a quantitatively demonstrated reduction in Ki-67-defined proliferative activity. Similar antitumour effects of nanoparticle-based carboplatin formulations have been reported in preclinical models.³³ However, the observed difference in tumour size should be interpreted as preliminary evidence of a therapeutic difference rather than proof of improved drug delivery or a specific mechanism. Although tumour size at the end of the experiment was significantly smaller in the nanomicelle-treated group than in the free-carboplatin group (p = 0.03), the longitudinal analysis of cumulative tumour burden using the area under the tumour-size–time curve did not show a significant between-group difference (p = 0.67). Therefore, the endpoint and longitudinal analyses should be considered complementary, and the present data do not establish a consistent therapeutic advantage of the nanomicellar formulation over free carboplatin across the entire observation period. In particular, the in vivo experiment used a subcutaneous tumour model and subcutaneous administration near the tumour, although the formulation is intended for intraperitoneal delivery. Therefore, the current experiment does not directly establish the behaviour or therapeutic efficacy of the nanomicelles after intraperitoneal administration. Validation in an intraperitoneal ovarian cancer model will be necessary to determine whether the observed antitumour effect translates to the intended route of administration.

Longitudinal monitoring of body weight provided no evidence of a marked difference in overall treatment tolerability between the groups, as body-weight changes remained limited and did not differ significantly between mice receiving free carboplatin and those receiving carboplatin-loaded nanomicelles (Figure S3). These findings should be interpreted cautiously, as body-weight monitoring alone provides only a limited assessment of treatment tolerability and does not constitute a comprehensive evaluation of systemic toxicity.

Histopathological examination suggested differences in the frequency of mitotic figures between the treatment groups. Ki-67, a widely used marker of tumour cell proliferation that has also been evaluated as a proliferation index in ovarian cancer,³⁴ was additionally assessed using a semiquantitative immunoreactive score (IRS). The field-level IRS data are presented descriptively and should be interpreted as an exploratory assessment rather than evidence of a quantitatively demonstrated difference in tumour cell proliferation between the treatment groups.

Several limitations should be considered when interpreting these findings. The formulation was developed using an exploratory approach rather than systematic optimization of the polymer-to-drug ratio, crosslinking conditions, and other processing parameters. Furthermore, the WST-1 assay was performed exclusively using SK-OV-3 ovarian cancer cells and did not include free carboplatin or unloaded nanomicelles as separate comparator groups. Therefore, the findings provide only a preliminary indication of in vitro cytocompatibility and do not allow the specific contribution of carboplatin encapsulation to the observed cellular response to be determined. In addition, the limited number of independent experimental runs, particularly for the 72-h exposure period (n = 2), restricts the strength of statistical inference regarding concentration- and time-dependent effects in the WST-1 assay. Future in vitro studies should include appropriate non-malignant cells as well as free carboplatin and unloaded nanomicelles as comparator groups. Although Ki-67 immunoreactivity was assessed semiquantitatively using IRS, the available field-level data could not be linked to individual tumour samples for analysis at the level of independent biological replicates. Therefore, no between-group inferential statistical analysis of Ki-67 IRS was performed. Further studies using a predefined quantitative labeling index or digital image analysis, with analysis performed at the level of independent tumour samples, would provide a more robust assessment of tumour proliferative activity. Moreover, the available body-weight data provide only a limited assessment of treatment tolerability and should not be interpreted as a comprehensive evaluation of systemic toxicity. In addition, direct studies of carrier residence, drug concentration in the peritoneal cavity over time, and pharmacokinetic behaviour were not performed. These limitations are particularly relevant because the intended application is intraperitoneal delivery, whereas the current in vivo efficacy study used a subcutaneous tumour model.

Within these limitations, the present study provides a proof-of-concept for hydrophobically modified chitosan as a carboplatin carrier. The formulation combines carboplatin encapsulation, preliminary in vitro cytocompatibility in SK-OV-3 cells, cellular internalization, and a preliminary antitumour effect with physicochemical properties that warrant further optimization. Further studies should focus on formulation optimization, validated long-term release measurements, direct assessment of peritoneal residence and drug exposure, quantitative analysis of cellular uptake, and evaluation in an appropriate intraperitoneal ovarian cancer model. These experiments will determine whether the physicochemical and biological properties observed in the present study translate into a measurable therapeutic advantage over free carboplatin.

Conclusions

Hydrophobically modified chitosan based nanomicelles represent a feasible carrier platform for carboplatin and warrant further investigation for localized drug delivery in ovarian cancer. The main advantages of the developed system include the use of a mucoadhesive chitosan-based carrier, the ability to incorporate carboplatin, and microscopically demonstrated cellular internalization. However, the present study should be regarded as a proof-of-concept, and the therapeutic advantage of the nanomicellar formulation requires further confirmation.

Future studies should focus on systematic formulation optimization, comprehensive characterization of drug release and long-term stability, quantitative evaluation of cellular uptake and tumour proliferation, and validation in an intraperitoneal ovarian cancer model. These investigations will be essential to determine whether the proposed nanocarrier can provide a meaningful therapeutic advantage over free carboplatin and support its further translational development.

Funding Statement

The paper publication was financially supported by the Minister of Science under the “Regional Initiative of Excellence” (RID) programme (RID_1_2025_D1_14).

Disclosure

The authors report no conflicts of interest in this work.