Pulsed Ultrasound-Mediated Drug Delivery Enhancement Through Human Sclera

Abstract

Purpose: The purpose of this study was to characterize whether pulsed ultrasound (PUS) affects transscleral drug delivery.

Methods: Fluorescein sodium (NaF, 376 Da) and fluorescein isothiocyanate-conjugated dextran 40 (FD-40, 40 kDa) were used as model drugs. Human sclera grafts were placed in modified Franz diffusion cells and were treated by PUS (1 megahertz [MHz], 0.71 W/cm², duty cycle 30%, application time 5 minutes) once or repeatedly under various conditions to assess permeation enhancement and reservoir effect. The safety of PUS application was assessed on human sclera grafts ex vivo and rabbit eyes in vivo by histology and temperature measurements.

Results: Single PUS application yielded a significant increase in FD-40 permeation (P <  0.05). Repeated PUS applications led to a further enhancement in FD-40 permeation and also significantly promoted NaF permeation (more than 8.51-fold, P <  0.05). The human scleral permeability was temporarily modified by PUS, as evidenced by the increased scleral permeability during PUS application and the unchanged permeability coefficients at steady state. The reservoir effect of human sclera was also enhanced by PUS application. Cavitation was detected under PUS. A minor increase in graft temperature rise (<1°C) and no ocular damage was caused by PUS.

Conclusions: PUS is an efficient and safe method to enhance model drugs to transport across human sclera by increasing the scleral permeability transiently and improving the reservoir effect. The enhancement was correlated with the molecule size and further promoted by the repeated PUS application.

Translational Relevance: Our study provides proof of concept for using PUS to enhance drug delivery to the posterior eye segment.

Introduction

Retinal and choroidal diseases are leading causes of irreversible vision impairment.^1,2 Treatment primarily involves intravitreal antivascular endothelial growth factor agents for neovascularization and corticosteroids for inflammatory conditions.^3,4 However, these invasive treatments carry risks, including retinal detachment, cataracts, and endophthalmitis,^5,6 underscoring the need for noninvasive, efficacious drug delivery alternatives. 

Transscleral permeation offers a viable pathway for drugs to reach therapeutic levels within the eye, as evidenced by sub-tenon administration of corticosteroids for uveitis.^7,8 However, transscleral delivery of large molecules, such as monoclonal antibodies, remains challenging because sclera is not the only barrier involved and those drugs have to pass through static, dynamic, and metabolic barriers before reaching the posterior segment.⁹ To overcome this difficulty, several optimized strategies have been proposed, including chemical penetration enhancers, photodynamic enhancing, microneedles, iontophoresis, and ultrasound application.^10–12 

Ultrasound, a diagnostic and therapeutic tool widely applied in clinical practice, has recently been recognized as an effective method for enhancing drug delivery based on its physical properties creating thermal effect, cavitation, and acoustic streaming.^13–16 Most studies use continuous ultrasound (100% duty cycle) for transscleral drug delivery.^17–21 Compared to continuous ultrasound, pulsed ultrasound (PUS) offers advantages in safety²² and, to some extent, is more effective in enhancing tissue permeability.²³ However, to our knowledge, ultrasound in pulsed mode has not been investigated in drug delivery through the human sclera. This study aims to investigate the efficacy and safety of PUS in enhancing the permeation of drugs of both small and large molecular weights across human sclera ex vivo, as well as to identify its potential mechanisms. The ultrasound parameters were meticulously selected. We utilized a frequency of 1 megahertz (MHz), informed by prior studies that have established its effectiveness in rabbit sclera.^17,18 Based on the duty cycle ranges from previous studies^23,24 and our preliminary experiments, 30% duty cycle was chosen for its balance of permeability enhancement and tissue safety. 

Materials and Methods

Reagents, Tissues, and Animals

Fluorescein sodium (NaF, MW 376 Da; Sigma-Aldrich Corp., St. Louis, MO, USA) and fluorescein isothiocyanate-conjugated dextran 40 (FD-40, MW 40 kDa; Sigma-Aldrich Corp., St. Louis, MO, USA) were used as small- and large-molecule model drugs, and were dissolved in phosphate-buffered saline (PBS) solution to produce final concentrations of 0.03 mg/mL and 0.1 mg/mL, respectively. The solution was protected from light by using aluminum foil until fluorescence-related assays described below. 

The use of human cadaver tissues was approved by the Human Research and Ethics Committees of the Eye and ENT Hospital of Fudan University in Shanghai, China. Thirty-six human cadaver donor eyes, including both men and women, with no history of high myopia, glaucoma, and eye diseases other than cataract were obtained from the eye bank at Eye and ENT Hospital of Fudan University within 24 hours after their death. The mean donor age at the time of death was 69.90 ± 5.59 years. The donor eyes were stored in moist chambers at 2 to 4°C for 2 ± 0.85 days to keep the tissue viability, as previously described.²⁵ 

All use of animals was approved by the Animal Research and Ethics Committees of the Eye and ENT Hospital of Fudan University in Shanghai, China. Six healthy New Zealand white rabbits (2–3 kg, 4–6 months old, male, clean grade) were purchased from Yingen Biotechnology Co., Ltd. 

Ultrasound Device and Its Calibration

The ultrasound parameters were carefully selected to optimize safety and efficacy. A frequency of 1 MHz and a duty cycle of 30% were chosen based on comprehensive review of the literature and a series of preliminary experiments that evaluated duty cycles of 30%, 40%, and 50%. All tested duty cycles effectively promoted drug permeation through the sclera without causing damage. However, the 30% duty cycle resulted in the least elevation of scleral temperature, enhancing tissue safety. A portable ultrasonic physiotherapy device, UT 1032 (Nu-Tek, Hong Kong, China), was used in this study. It provides ultrasonic energy at a frequency of 1 MHz and has a radiation area of 8.548 cm². The center frequency and the ultrasound intensity were calibrated according to Chinese National Standard GB/T 7965-2002. The measurements were made using a needle hydrophone (NH 0200 Hydrophone; Precision Acoustics, Dorchester, Dorset, UK) with a measured sensitivity of 39 mV/MPa at 1 MHz. Specifically, the ultrasound device was set to produce ultrasound at a frequency of 1 MHz, an ultrasound intensity of 0.72 W/cm², and a duty cycle of 30%, which was measured by the hydrophone. The hydrophone was situated 200 mm distant from the ultrasound transducer probe and attached to an oscilloscope (MSO5102; RIGOL, Beijing, China). It was determined that the ultrasound's center frequency, duty cycle, and intensity were 1 MHz, 30%, and 0.71 W/cm², respectively (see Supplementary 1 for the calculation of ultrasound intensity). 

Human Sclera Graft Preparation

Four round scleral areas, 6 mm posterior to the limbus, were chosen and cut into grafts. Extraocular muscles, perforating ciliary vessels, and vortex veins were avoided during grafting. The grafts were sandwiched between two glass slides and a digital caliper (resolution 0.01 mm; Chengdu Chengliang Chuanpai, Chengdu, China) was used to measure their thickness. The mean thickness of the isolated sclera was 0.69 ± 0.08 mm. This procedure normally took 10 to 15 minutes. The grafts were then rinsed with and equilibrated in PBS at room temperature for 25 minutes. 

Ex Vivo Experiment’s Setup

Each human sclera graft was mounted on a modified Franz diffusion cell (effective diffusion area, 0.5024 cm²) with the orbital side facing the donor chamber, and was clamped by a screw to prevent lateral leakage and tissue edge abrasion. The donor chamber contained 1 mL of model drugs (either NaF or FD-40). The receiver chamber was filled with 5 mL of degassed PBS (pH 7.4), thermostated at 37°C, and stirred with a magnetic stirrer (200 rpm) to avoid the boundary layer effect. The ultrasound transducer probe was immersed in the model drug solutions and was positioned approximately 15 mm above the surface of the human sclera, as it was the location for the maximum pressure of the 1 MHz ultrasound (Fig. 1). The ultrasound device was set to produce an ultrasound with a frequency of 1 MHz, an intensity of 0.71 W/cm², a duty cycle of 30%, a pulse repetition frequency (PRF) of 100 Hz, and a pulse duration of 3 milliseconds (ms), and the application time was 5 minutes. No ultrasound absorber was placed in the modified Franz diffusion cell. 

Figure 1.

 

Schematic diagram of the ex vivo setup. An isolated human sclera was placed between the donor and receiver chamber of the modified-Franz diffusion cell with the orbital side facing the donor chamber. The receiver chamber was filled with PBS, and the donor chamber was filled with model drug solutions. The ultrasound transducer probe was positioned 15 mm above the human sclera.

Transscleral Permeation Experiments

To assess drug permeation enhancement immediately after PUS application, the human sclera was immediately removed from the diffusion cell for cryo-sectioning after a 5-minute experimental exposure (Fig. 2A). 

Figure 2.

 

Schematic diagram of the experimental procedure. (A) The human sclera was immediately removed from the diffusion cell for cryo-sectioning after a 5-minute experimental exposure. (B) In permeation experiments, PUS was applied to human sclera once to 3 times with an interval of 5 minutes, and receiver chamber solution was sampled at predetermined time-points (0.25, 0.5, 0.75, 1, and 2 hours for NaF, and 0.25, 0.5, 0.75, 1, 2, 3, and 4 hours for FD-40). (C) In the PUS pretreat test, the human sclera graft was first exposed to PBS for 25 minutes with concurrent PUS application three times, and then was exposed to the model drugs. (D) PUS was applied three times, then both the model drugs and ultrasound were removed, and the sampling of the receptor solution was continued up to 2 hours and 4 hours for NaF and FD-40, respectively. PUS, pulsed ultrasound; 1 × PUS, once PUS application; 2 × PUS, twice PUS applications; 3 × PUS, 3 times PUS applications.

In addition, to characterize a time-dependent drug permeation enhancement, the duration of permeation experiments was set to 2 hours for NaF and 4 hours for FD-40, respectively. In the ultrasound-treated group, PUS was applied to human sclera 1 to 3 times with an interval of 5 minutes. After the PUS was stopped, the grafts were kept in contact with the model drugs until the proposed permeation duration was achieved. Receiver chamber solution was sampled at predetermined time points (0.25, 0.5, 0.75, 1, and 2 hours for NaF, and 0.25, 0.5, 0.75, 1, 2, 3, and 4 hours for FD-40) and simultaneously replaced with an equal volume of fresh PBS. The control group did not receive PUS treatment (Fig. 2B). 

Moreover, to evaluate the alterations of scleral permeability, a PUS pretreatment test was performed. The human sclera graft was first exposed to PBS for 25 minutes with concurrent PUS application 3 times. Subsequently, PBS was replaced with the model drug solution, and the sclera was exposed to the model drug. For NaF, receptor solution was collected at 0.25 hours for quantification. For FD-40, the sclera was removed after 5 minutes for cryo-sectioning (Fig. 2C). 

Additional tests assessed the potential role of the sclera as a drug reservoir. PUS was applied 3 times, then both the model drugs and ultrasound were removed, and receptor solution sampling continued for 2 hours for NaF and 4 hours for FD-40 (Fig. 2D). 

Transscleral Penetration Distance and Average Fluorescence Intensity Measurements

The human sclera was removed from the diffusion cell and immersed in optimum cutting temperature medium (OCT; Sakura, Torrance, CA, USA) and quickly frozen at –80°C for cryo-sectioning. The sclera was cryo-sectioned into 10-µm thick sections using a cryostat (CM1850; Leica, Wetzlar, Germany). 

All sections were imaged under a fluorescence microscope (Leica Microsystems, Bensheim, Germany) with the same parameter settings (bandpass filter = 460–490 nm, dichroic mirror = 505 nm, barrier filter = 515 nm, and exposure time = 1000 ms). Penetration distance and average fluorescence intensity were measured and quantified using ImageJ software (version 1.53e; http://imagej.nih.gov/ij; National Institutes of Health, Bethesda, MD, USA). 

The penetration distance was measured according to the methods reported by Suen et al.²⁶ Briefly, the distance from the scleral surface to the depth where the fluorescent signal fell to the statistical threshold represented the penetration distance. The statistical threshold (i.e. background signal of the slide) was defined as the average signal of the sclera not exposed to model drugs. Average fluorescence intensity was quantified by changing the images to greyscale. At least 20 measurements were performed in different regions of each diffusion area, including the peripheral and central sclera. 

To ensure measurement consistency and repeatability, we conducted repeatability and reproducibility analysis. Repeatability and reproducibility were assessed in a random subset of 15 images. A single examiner (author S.Q.Y.) measured each image twice on different days to decide intra-examiner variability. Additionally, a second examiner (author W.C.) measured the same images independently on a different day to decide inter-examiner variability. The coefficient of the intra-class correlation (ICC) was used to calculate the intra-examiner and inter-examiner variability. 

Model Drug Quantification and Data Processing

Sample concentrations were quantified using a fluorometer. Standard solutions were prepared and analyzed with a microplate reader (Tecan, Mannedorf, Switzerland) to make standard curves of fluorescence versus concentration. Wavelengths for excitation and emission for NaF and FD-40 were 460 and 585 nm, and 515 and 535 nm, respectively. 

The cumulative amount of model drugs permeated per unit area of sclera (Q, mg/cm²) was plotted against time (t, h). The transscleral flux (J_s, mg/cm²h) at steady state was calculated from the slope of the linear portion of the plot (ΔQ/Δt). The transport lag time (t_lag) was the x-intercept of the linear regression line in the plot. The permeability coefficient (K_p, cm/h) was defined as the flux normalized by the donor concentration (C₀, mg/mL) ²⁷:  

\begin{equation}{K_P} = \frac{{\Delta Q}}{{\Delta t{C_0}}}\end{equation}

(1)

 

This formula is only applicable if the cumulative total amount of model drugs transferred to the receiver chamber is less than 5% of the total amount in the donor chamber. All the values in this study were under this limit. 

Detection of Cavitation by Acoustic Emission

To investigate whether cavitation existed under the proposed ultrasound conditions, the passive cavitation detection method was used to measure acoustic bubbles activity with a needle hydrophone (NH 4000 Hydrophone; Precision Acoustics, Dorchester, Dorset, UK) with a sensitivity of 1000 to 1200 mV/MPa at the transducer driving frequency of 1 to 5 MHz. The ultrasound transducer and the hydrophone were positioned in a water tank filled with ultra-pure water produced by the Millipore Milli-Q purification system (Millipore Corp., Bedford, MA, USA). The hydrophone was fixed with its tip 60 mm from the center of the ultrasound transducer probe and was connected to an oscilloscope (MSO5102; RIGOL, Beijing, China). When the PUS was on, the voltage signal measured by the hydrophone in response to the ultrasound was recorded by the oscilloscope and was transferred to a computer to set to Fast-Fourier-Transform (FFT) mode to display the frequency spectrum by MATLAB (MATLAB, The MathWorks, Natick, MA, USA). 

Safety Assessment

To evaluate the potential thermal effects of PUS on the sclera, temperatures on the scleral graft surface before and after PUS application were monitored ex vivo using an infrared thermometer (KM-520; GAOMU, Shenzhen, China). Temperature measurements were taken at four time points: before the PUS application, and immediately after the first, second, and third PUS applications. The temperature rise was calculated by subtracting the baseline temperature (before PUS) from the temperatures recorded after each PUS application. 

The safety of the PUS application on the sclera was assessed ex vivo and in vivo by identifying potential morphological and ultrastructural changes of ocular tissues via histological and transmission electron microscopy (TEM) assays. 

The setup was identical to the aforementioned in “Ex Vivo Experiment's Setup” except that 1 mL of PBS solution, instead of the model drugs, was placed in the donor chamber. The PUS was applied to human sclera 3 times with a 5-minute interval. After then, the human sclera was immediately removed from the setup, fixed in 4% paraformaldehyde, and immersed in 4% paraformaldehyde for 24 hours. Then, they were dehydrated with graded series of alcohol, paraffin-sectioned with a 5-µm thickness using a microtome (RM2235; Leica, Hesse, Germany), and stained with hematoxylin and eosin (H&E) and Sirius-red for light microscopy (Leica Microsystems, Bensheim, Germany) and polarized light microscopy (Nikon Co, Tokyo, Japan), respectively. Sirius-red, which binds parallel to collagen molecules, was used to enhance collagen birefringence. To analyze the scleral organization quantitatively, collagen bundle proportionate area, and collagen bundle orientation in H&E-stained images were assessed using ImageJ software. 

To further evaluate the PUS-associated structural alteration of collagen bundles, TEM observations were performed. The sclera was sampled and fixed in 2.5% glutaraldehyde at 4°C for 24 hours, followed by a 2-hour incubation in 1% osmium. Then, the tissue was dehydrated in a graded ethanol series and embedded in an epoxy resin mixture at 60°C for 48 hours. Thin sections were cut using a microtome and photographed on a TEM (JEM-1230EX; JEOL, Tokyo, Japan). 

In addition, to identify whether PUS application morphologically injured the choroid and retina, we performed in vivo tests on rabbits. New Zealand white rabbits were anesthetized by intramuscular injection of 10% xylazine hydrochloride (0.1 mL/kg) and intravenous injection of 1.5% sodium pentobarbital (1 mL/kg). An eye cup containing 1 mL of PBS was placed on the superior temporal area of the sclera (1.5 mm to the limbus), and the ultrasound transducer probe was positioned approximately 15 mm above the sclera and immersed in PBS. Then, the tested eyes were exposed to PBS and three PUS applications. Immediately after the PUS application, the rabbits were euthanized, and the application site was marked, followed by the enucleation of the whole eyeball. The eyes were then fixed, embedded, cut, stained with H&E, and observed as previously described. 

Statistical Analysis

All values are presented as mean ± standard deviation (SD). Statistical differences between groups were determined by the Mann-Whitney test. Differences were considered statistically significant at a level of P <  0.05. 

Results

Increased FD-40 Penetration in Human Sclera Under PUS

The intra-examiner ICCs for penetration distance and fluorescence intensity were 0.996 and 0.981, respectively, whereas the inter-examiner ICCs were 0.971 and 0.924, indicating excellent measurement repeatability and reproducibility. 

In the control group (n = 6), without PUS treatment, FD-40 penetrated 52.64 ± 16.37 µm into the sclera, with an average fluorescence intensity of 58.71 ± 6.87 AU. Following a single PUS application (n = 6), the penetration distance of FD-40 increased by 3.06-fold to 161.32 ± 35.35 µm, and the average fluorescence intensity increased by 1.45-fold to 84.88 ± 12.91 AU (P <  0.05). These results demonstrate that PUS significantly enhances the penetration of FD-40 through human sclera (Figs. 3A–C). 

Figure 3.

 

(A–C) Representative images of cryo-sectioned human sclera exposed to FD-40 in the absence and presence of PUS application, and the penetration distance of FD-40. (D–F) Representative images of cryo-sectioned human sclera exposed to NaF in the absence and presence of PUS application, and the average fluorescence intensity of model drugs. C, choroidal side of the sclera; O, orbital side of the sclera; scale bar = 100 µm; *P <  0.05.

In NaF, full-thickness fluorescence was observed in both control (n = 6) and PUS-treated samples (n = 6; Figs. 3D, 3E). The difference in average fluorescence intensity between the PUS-treated group and the control group was not significant (93.13 ± 21.97 AU vs. 79.32 ± 20.63 AU, P > 0.05; Fig. 3F). 

PUS-Mediated Transscleral Drug Permeation Enhancement

Permeation profiles of model drugs across human sclera (Fig. 4) demonstrated the ability of PUS to enhance transscleral drug permeation (see Supplementary 2 and 3 for the data of transscleral drug permeation experiments). Table 1 summarizes the permeability coefficients and lag times of model drugs across the human sclera. 

Figure 4.

 

Permeation profiles of NaF (A) and FD-40 (B) in the absence and presence of PUS application. The receiver chamber solution was sampled at predetermined time points (0.25, 0.5, 0.75, 1, and 2 hours for NaF, and 0.25, 0.5, 0.75, 1, 2, 3, and 4 hours for FD-40). In the PUS-treated group, PUS was applied to human sclera once to 3 times with an interval of 5 minutes. Control, control groups; 1 × PUS, once PUS application; 2 × PUS, twice PUS applications; 3 × PUS, 3 times PUS applications. (A) *P <  0.05 = 2 × PUS versus control; ^†P <  0.05 = 3 × PUS versus control. (B) *P <  0.05 = 1 × PUS versus control; ^†P <  0.05 = 2 × PUS versus control; ^‡P <  0.05 = 3 × PUS versus control.

Table 1.

 

Permeability Coefficients and Lag Times of Model Drugs Across Human Sclera

 

For NaF, a single PUS application did not result in significant differences in the permeation profile compared to controls (see Fig. 4A). However, after 2 PUS applications, an 8.51-fold increase in NaF permeation at 0.25 hours was observed (P <  0.05), although subsequent time points and the permeability coefficient were not significantly different from the controls (P > 0.05). Three PUS applications further enhanced NaF permeation up to 1 hour (P <  0.05), but the permeability coefficients remained similar to those of the controls (P > 0.05; see Table 1). Additionally, the lag times for NaF in both control and PUS-treated groups showed no significant differences (see Table 1). 

The PUS-mediated permeation enhancement of FD-40 was more pronounced than that of NaF. As shown in Figure 4B, a single PUS application resulted in detectable FD-40 permeation at 0.25 hours (n = 6), whereas FD-40 was not detected until 0.75 hours in the control group (n = 10). Repeated PUS applications further enhanced FD-40 permeation (two applications: n = 9 and three applications: n = 8). The duration of permeation enhancement was 1 hour for a single application and 2 hours for repeated applications (see Fig. 4B). Repeated PUS applications significantly shortened the lag times of FD-40 (see Table 1; P <  0.05), whereas the permeability coefficients remained statistically unchanged (P > 0.05). 

PUS-Modulated Scleral Permeability Changes

PUS pretreatment was applied to human sclera to evaluate changes in scleral permeability. If human scleral permeability cannot be increased by PUS or can only be increased but then immediately returns to normality when the PUS stops, the transscleral permeation in the PUS-pretreated group is expected to resemble the control without PUS pretreatment. A 9.38-fold increase in NaF permeation was observed in the PUS-pretreated group (n = 6) compared to the control group (n = 6) at 0.25 hours, with values of (2.46 ± 0.89) × 10⁻⁵ mg/cm² and (2.62 ± 5.03) × 10⁻⁶ mg/cm², respectively (P <  0.05). 

Figure 5 demonstrates the transscleral penetration distance of FD-40 (Fig. 5A is identical to Fig. 3A) in the PUS-pretreated group (123.2 ± 38.24 µm) was 2.34 times longer than in the control group (52.64 ± 16.37 µm, P <  0.05; Fig. 5A–C). However, average fluorescence intensity of FD-40 was not significantly different between the PUS-pretreated (59.05 ± 7.82 AU, n = 6) and the control groups (58.71 ± 6.87 AU, n = 6; Fig. 5D). 

Figure 5.

 

Representative images of cryo-sectioned human sclera exposed to FD-40 in the absence (A) and presence (B) of PUS pretreatment, and the penetration distance (C) and average fluorescence intensity (D) of FD-40 measured by Image J. C, choroidal side of the sclera; O, orbital side of the sclera; white line in (A) and (B), penetration distance of FD-40; control, control groups; PUS, PUS-pretreated groups; scale bar = 100 µm; *P <  0.05.

Reservoir Effect Alterations Caused by PUS

To assess the reservoir effect of human sclera altered by PUS, both model drugs and ultrasound device were removed after exposed to human sclera for 25 minutes. Permeation was followed up to 2 hours and 4 hours for NaF and FD-40, respectively. Figure 6 shows significant differences in permeation profiles between the PUS-treated and the control groups (P <  0.05). The amount of drug permeated from 25 minutes to the final time point reflects the drug accumulated in the sclera during the initial 25 minutes. 

Figure 6.

 

Post-PUS transport of NaF (A) and FD-40 (B). The PUS was applied three times (0–25 minutes). After the PUS ended, the donor solution (NaF or FD-40) was removed. The receiver chamber solution was sampled at predetermined time points (0.25, 0.5, 0.75, 1, and 2 hours for NaF, and 0.25, 0.5, 0.75, 1, 2, 3, and 4 hours for FD-40). Control, control groups; PUS, three times PUS applications. *P <  0.05 PUS versus control.

NaF permeation from 25 minutes to 2 hours in the PUS-treated group (n = 6) was 1.23 times higher compared with the controls (n = 6, 9.21 ± 1.91 × 10⁻⁴ mg/cm² vs. 7.47 ± 0.99 × 10⁻⁴ mg/cm², P <  0.05). FD-40 permeation from 25 minutes to 4 hours in the PUS-treated group (n = 6) was 1.66 times that of the controls (n = 6, 7.45 ± 1.65 × 10⁻³ mg/cm² vs. 4.48 ± 0.74 × 10⁻³ mg/cm², P <  0.05). 

Evaluation of Cavitation Activity

Peaks at harmonic frequencies are indicators of cavitation.²⁸ In our study, when the PUS was applied, peaks at the fundamental frequency (f, 1 MHz) and harmonic frequencies (2f and 3f) on the frequency spectrum were identified, suggesting the presence of cavitation (Fig. 7). The control group, without PUS treatment, was not expected to show cavitation; thus, their harmonic frequency analysis was omitted. 

Figure 7.

 

Frequency spectrums for PUS in the range of 0.2 to 1.2 MHz (A) and 0 to 5 MHz (B). f, peak at driving frequency; 2f and 3f, peaks at harmonic frequencies; f/2, peak at subharmonic frequency.

Safety Assessment

Morphological and ultrastructural examinations of human sclera and rabbit retinal tissues post-PUS treatment showed no structural alterations compared with the controls (Fig. 8). Quantitatively, the collagen bundle proportionate area and the collagen bundle orientation were analyzed in H&E-stained images. No significant differences were found between PUS-treated (n = 6) and untreated (n = 6) tissues: collagen bundle proportionate area (0.87 ± 0.04 vs. 0.83 ± 0.03, P > 0.05), and collagen bundle orientation (4.47 degrees ± 0.60 degrees vs. 4.80 degrees ± 0.65 degrees, P > 0.05). These quantitative findings corroborate the qualitative observations, indicating that PUS does not compromise the structural integrity of scleral tissues (see Figs. 8A, 8E). Sirius-red staining indicated well-oriented collagen bundles forming a regular network structure post-PUS treatment (see Fig. 8F). Initial observations indicate a similar ultrastructure of collagen fibrils between PUS-treated and control samples, with no apparent disorganization in PUS-treated samples upon qualitative assessment (see Figs. 8C, 8G). 

Figure 8.

 

Representative H&E staining (A, D, E, H), Sirius-red staining (B, F), and TEM (C, G) images of the human sclera (A–C, E–F) and rabbit retina (D, H). Postmortem without treatment (A–D); after three times PUS applications (E–H). GCL, ganglion cell layer; ILM, inner limiting membrane; INL, inner nuclear layer; IPL, inner plexiform layer; ONL, outer nuclear layer; OPL, outer plexiform layer; PR, photoreceptors; scale bar = A, B, D, E, F, and H = 50 µm; C, G = 0.2 µm.

Rabbit eyes after three PUS applications demonstrated no sign of retinal morphological abnormalities, such as deformations, degeneration, or inflammation (see Fig. 8H). 

The scleral temperature before PUS application was 30.43 ± 1.15°C. As shown in Table 2, the average PUS-associated temperature rises on the scleral surface were all less than 1°C (n = 6). 

Table 2.

 

Temperature Alteration of the Scleral Surface After PUS Application

 

Discussion

This study is the first to assess the use of PUS to enhance the delivery of small and large molecular weight drugs through human sclera. Although there have been studies utilizing ultrasound to facilitate drug delivery through rabbit and porcine sclera,^{17–21,26,29,30} the knowledge acquired from animal sclera cannot be directly applied to human sclera due to their different features, such as thickness, collagen fibril arrangement, and melanin content.^31–33 Using NaF and FD-40 as stand-ins for small molecules like dexamethasone (392 Da) and macromolecules like ranibizumab (48 kDa), this study aims to fill the gap in knowledge regarding PUS-facilitated drug delivery in human ocular treatment. 

Previous studies attempted to apply ultrasound in continuous mode to transscleral drug delivery.^{17–19,21,26,29,34} However, the continuous ultrasound could result in a dramatic temperature increase in the ocular tissue and hence thermal damage to the eye.^21,34 This problem could be circumvented by applying ultrasound in pulsed mode.²² Therefore, the PUS was applied here. A frequency of 1 MHz and a 30% duty cycle were chosen for their proven efficacy and favorable safety profile in our preliminary experiments. The application time of the PUS was set to 5 minutes, a duration informed by modeling studies indicating increased drug transport across biological barriers.³⁵ Results showed that PUS significantly enhances FD-40 penetration through the human sclera, as indicated by increased penetration distance and fluorescence intensity. This aligns with previous ex vivo studies utilizing continuous ultrasound for macromolecule delivery in rabbit sclera.^17,19 However, the PUS does not significantly affect NaF fluorescence intensity, suggesting the effect may be size-dependent. 

To further elucidate the impact of PUS on NaF penetration and to assess the temporal dynamics of PUS-mediated transscleral drug delivery, extended duration experiments were specifically designed. The duration of permeation experiments was set to 2 hours for NaF and 4 hours for FD-40, respectively, based on their molecular weights, diffusion characteristics, our preliminary studies, and established benchmarks in ocular pharmacokinetic studies, incorporating repeated PUS applications. The results indicate that PUS significantly enhances the permeation of NaF across human sclera, particularly with repeated applications. Whereas PUS accelerates initial drug permeation, it does not change the overall permeability coefficients or lag times, suggesting its primary effect is on the initial transport phase. Notably, PUS enhances FD-40 permeation more effectively than NaF. Repeated applications further amplify this effect, reducing lag times without altering permeability coefficients. This highlights PUS’s efficacy in enhancing transscleral delivery, especially for larger molecules like FD-40. 

Permeability coefficients for model drugs in PUS-treated groups and control groups were similar, indicating that PUS does not cause permanent changes to the human scleral structure. This is consistent with previous studies on rabbit sclera, where ultrasound was found to temporarily increase scleral permeability.^17,19 In our study, we extended this inquiry by applying PUS thrice to human sclera, observing a significant rise in drug permeation that stabilized thereafter, indicating a transient alteration in scleral permeability that normalizes over time. To evaluate the changes in scleral permeability, additional experiments used PUS to pretreat human sclera three times in our study, and the permeated amounts of the two model drugs were significantly increased with PUS pretreatment and then kept at a steady state, indicating a transient alteration in scleral permeability that normalizes over time. This differs from Huang et al.’s findings where a twofold increase in permeability was observed in rabbit sclera after a 5-minute ultrasound,¹⁸ likely due to different experimental conditions and species-specific responses. 

In addition, the reservoir effect of human sclera was identified in this study. In the control group, model drugs could still be detected in the receptor solution after model drugs were removed, due to the release of model drugs previously accumulated inside the human sclera to the receptor solution. In comparison, the PUS-treated group exhibited a greater release. These indicated that PUS application increased the scleral accumulation of drugs, enhancing the reservoir effect. It is speculated that the amount of drugs that penetrate through sclera increases with the enhanced scleral permeability and, meanwhile, those drugs could considerably accumulate into the sclera to form a drug reservoir. When the PUS application stops and scleral permeability recovers, drugs reserved in the sclera would release continually. This reservoir effect is of great importance in clinical practice, as it suggests that the PUS-enhanced drug delivery persists beyond the application period. 

Our study highlighted that the PUS treatment did not induce any significant structural changes in the human scleral or rabbit retinal tissues, as both morphological and ultrastructural examinations, along with quantitative analysis of collagen bundle proportionate area and orientation, showed no differences compared with the controls, confirming the preservation of tissue integrity post-treatment. This aligns with prior rabbit model outcomes.^26,29 Additionally, our previous in vivo study revealed no structural and functional alterations in the retina and choroid of rabbits 2 weeks after PUS treatment.³⁶ The sclera is an elastic and microporous tissue primarily composed of interlacing collagen fibrils and interfibrillar proteoglycans.³⁷ Proteoglycans that occupy the interfibrillar compartment play essential roles in solute diffusion and fluid movement through sclera.³⁷ Therefore, it is possible that ultrasound modifies proteoglycans temporarily and thus decreases the transport resistance of sclera. Previous studies suggested that altering the proteoglycan architecture could reduce the diffusional resistance against macromolecules.^19,38 These may also explain why the permeation enhancement of macromolecules was more remarkable than that of small molecules, as demonstrated in our study by the fact that a single PUS application significantly improved the permeation of FD-40 but had no effect on NaF permeation. Proteoglycan would gradually reassemble when the ultrasound application terminates, and the scleral permeability returns to its normal level. The restoration feature of the sclera is crucial, because the sclera is vital for maintaining the structural integrity and shape of the eyeball and protecting the eye. Further research is required to investigate the influence of ultrasound on proteoglycans at the ultrastructural level. 

The exact mechanism of PUS-mediated transscleral drug delivery enhancement remains to be fully understood, but it is likely related to the complex effects of ultrasound. Previous studies suggested that cavitation contributed to the enhancement of drug permeation across rabbit sclera.^17,19,29 Cavitation refers to the formation of bubble-like structures and their subsequent dynamics in a liquid medium in response to ultrasound,³⁹ and subsequently generates microstreaming and shear stress, or gives rise to shock waves and microjets. They could exert force on the adjacent cells and tissues and produce certain channels to promote drug delivery.⁴⁰ The generation of cavitation depends partly on ultrasound parameters.⁴⁰ According to previous studies, the currently used ultrasound parameters were in the range of cavitation generation.^{10,21,41–43} Our results demonstrated peaks at harmonic frequencies on the frequency spectrum, denoting the bubble activity from cavitation. Therefore, cavitation may contribute to transscleral drug delivery in this study. It is speculated that acoustic microstreaming in cavitation may introduce a chaotic flow to enhance the rate of drug transport and disturb scleral structure to modify scleral permeability during ultrasound application,^17,19 and then the scleral permeability gradually returns to a normal value after ultrasound application. In addition, the lack of an ultrasound absorber in the Franz diffusion cell poses a challenge in eliminating the impact of reflected waves in the experimental setup. It is probable that such reflected waves may have a role in altering the permeability of the sclera, necessitating further investigation to authenticate their influence. 

There is minimal temperature elevation after PUS application in the current study. This is also reported in a previous study that transscleral permeation enhancement is not due to thermal effect.³⁴ Human retinal pigment epithelium (RPE) cells were found to be tolerable to temperatures of 39.5 to 40°C due to high blood flow in the choroidal working as a heat sink.⁴⁴ Concerns for ultrasonic thermal injury on the human ocular tissues are further minimized by delivering in a pulsed mode.²² Our results showed that the scleral temperature change was minimal (all temperature elevations <1°C) and thus cannot be responsible for the permeation enhancement. 

There are several limitations in this study. The primary constraints include the use of a limited range of model compounds. Whereas NaF and FD-40 served as proxies for hydrophilic small and large molecules, respectively, they do not encompass the full spectrum of physicochemical properties of ophthalmic drugs, particularly regarding lipophilicity and charge. Future work should explore a variety of compounds with diverse lipophilicity and charge to better model the range of ophthalmic drugs. Additionally, the ex vivo nature of our experimentation cannot fully replicate the dynamic in vivo environment of the eye, particularly the active transport mechanisms within the RPE and the choroidal blood flow. Further research using in vivo models and clinical trials will be necessary to fully elucidate the influence of PUS on transscleral drug delivery. Moreover, further investigation into the bioeffects of ultrasound on scleral proteoglycans is required to understand the mechanism behind ultrasound-enhanced transscleral drug delivery and to refine the ultrasound parameters that optimize transscleral permeation. Although we affirm that current parameters are effective in this study, it is important to recognize that they may not be optimal for all situations. The optimal parameters may vary not only based on the specific drug being delivered but also in relation to the condition of the scleral tissue. Further research should continue to explore the optimal ultrasound settings. 

Conclusions

In summary, our study demonstrates that PUS significantly enhances drug delivery through the human sclera by temporarily increasing scleral permeability and augmenting the scleral reservoir effect, with the impact potentially being size-dependent. Repeated applications of the PUS notably improved transscleral drug delivery. Cavitation may play a role. Importantly, no structural alterations were observed in the human scleral or rabbit retinal tissues, and temperature increases were minimal. These findings suggest that PUS is an effective and safe method for enhancing transscleral drug delivery. 

Acknowledgments

The authors thank Yan Wang from the Office of Research, University of Michigan, for assistance with language editing and data evaluation, and Yuanhao Wu for his kind help during the research. 

Supported by National Natural Science Foundation of China (grant numbers 82471099, 11774382 and 82201175); Science and Technology Commission Shanghai Municipality (grant number 20S31905900); and the Shanghai Sailing Program (grant number 20YF1405100). 

Disclosure: S. You, None; S. Wu, None; S. Yang, None; Z. Zhao, None; W. Chen, None; X. Chen, None; H. Wang, None; Q. Xia, None; J. Xiong, None; H. Zhou, None; X. Mo, None