During the last years the imaging technologies, such as electron paramagnetic resonance spectroscopy, positron emission tomography or magnetic resonance imaging (MRI) have gained further importance for studying age-related changes of the volume, structure of eye tissues and drug distribution after periocular and intravitreal injection .
With advanced age the gelatinous fraction of the vitreous body decreases and liquid filled pockets are formed. This socalled posterior vitreous detachment (PVD) may cause the separation of the vitreous membrane from the retina . With increasing amount of liquefaction, the convective flow in the vitreous cavity induces increased drug mobility . As the vitreal liquefaction (VL) must be considered to investigate drug distribution in vivo, ex vivo and in vitro, the PVD was implemented in this study.
Beyond physicochemical features of the active agent, procedure of intravitreal injection , patient-age and progression of the PVD, the eye movement has a distinctive effect on the intravitreal distribution. Saccadic eye movements induce vitreous humor motion, determine intravitreal distribution processes and shear stresses along the surrounding membranes . The distinctive effect of the eye movement, particularly under PVD conditions, was shown in vitro with the combination of the eye movement system (EyeMoS) and the vitreous model (VM ). With increasing amount of liquid in the VM, the influence of the movement gained importance. The usage of MRI as visualization technology implies many benefits such as a higher spatial resolution and a continuous monitoring without sampling .
2 Material and methods
Solutions of gadolinium-based contrast agents with different molecular masses, GadoSpin™ M (GSM, 938 Da), GadoSpin™ D (GSD, 17000 Da) and GadoSpin™ P (GSP, 20000 Da) obtained from Miltenyi Biotec GmbH (Bergisch-Gladbach, Germany) were used to investigate distribution effects depending on molecular weight over a period of 12 h. Sterile 0.9 % sodium chloride solution (NaCl) was purchased from B. Braun Melsungen AG (Melsungen, Germany). Rotiphorese®, ammoniumperoxodisulfate (APS) and tetramethylethylenediamine (TEMED) were supplied from AppliChem GmbH (Darmstadt, Germany). All chemicals were of analytical grade.
Fresh porcine eyes were obtained as animal byproducts from a registered slaughter facility. The eyes were excised immediately after the death of the animal and stored in Ringer buffer maximally for 1 h until use.
2.1 Vitreous model and eye movement system
The previously developed combination of the VM (Figure 1) and the EyeMoS allows simulation of the natural vitreous body in form, shape, selected physicochemical properties as well as several types of eye movement in vitro .
To achieve closer conformity with the situation in vivo and to simulate the progress of PVD , additional experiments were performed by substituting 50 % of PAA-gel by Ringer buffer solution. After completely gelatinization the PAA-gel was overlaid with Ringer buffer solution. Through implementation of a second vent, the test solution was injected either through the gelatinous fraction or through the liquid phase to investigate the relevance of the injection procedure . For distribution experiments the VM was mounted on the holder of the EyeMoS.
A periodic movement was accomplished for 6 h to investigate the initial drug distribution. The movement was interrupted for a period of measurement in the MRI device. To compare the distribution in VM with a commonly consulted animal model , contrast agents were injected into porcine vitreous (ex vivo).
2.2 MRI distribution studies
After completely gelatinization of PAA-gel 100 µL of the dissolved contrast agent were injected centrally into the PAAgel of the VM using a 22 gauge needle. Injection into the vitreous body of the porcine eye was performed lateral to the cornea through the sclera, following usual surgical technique. For injection into the VM the syringe was placed on a spacer on top of the vent to ensure the reproducibility of needle penetration depth.
Distribution in the model was investigated using a 7.1 Tesla magnetic resonance imaging device (Clin-Scan 70/30, Bruker Bio-Scan GmbH, Ettlingen, Germany) with a phased array surface coil (rat brain, Bruker Bio-Scan GmbH, Ettlingen, Germany) with 2 channels and 2 coil elements for each channel (T1weighted sequences, T1w). The applied MRI parameters are listed in Table 1. The acquisition time of a single measurement amounted to 8:02 minutes.
Parameters for magnetic resonance imaging.
|Field of view||40 mm x 40 mm|
|Matrix||130 pixels x 130 pixels|
|Spatial resolution||125 µm x 125 µm|
|Slice thickness||700 µm|
|Flip angle||90 °|
The analysis of the obtained images was accomplished with OsiriX TM imaging software . By setting the range of signal intensities, the regions of interest and the resulting volumes of the injected contrast solution were tracked and calculated by the software. For the calculation of the volume, all determined slices of 700 µm thickness were included. The scale of the volume expansion in the VM or porcine vitreous body was determined by setting the volume short after injection as point of origin. The ratio between the start value and the volumes at the selected time points were calculated. All experiments were performed in triplicate.
3.1 Distribution of contrast agents with different molecular weights
The molecular weight of the selected contrast agents corresponds broadly to those of therapeutically applied active agents. To estimate the distribution of clinical established drugs with low molecular weight as dexamethasone  or triamcinolone acetonide , dissolved GSM was used. GSD and GSP were applied to simulate the distribution of macromolecules such as antibodies (e.g. bevacizumab, 150 kDa) and monoclonal antibody fragments (e.g. rituximab, 40 kDa ). Slow pursuit eye movements were simulated .
As expected, with increasing molecular weight of injected contrast agent, the velocity of distribution decreased (Figure 2). Simultaneously the spatial volume of distribution in VM and in porcine vitreous humor decreased.
Following central injection into simulated native state of the VM (100% PAA-gel), hyperintense wavy injection tracks of the contrast agents were detected on T1w-images (Figure 2, t=0 h). Independent of the molecular weight, an initial backflow along the puncture duct was seen on MR images in VM as well as in enucleated porcine eyes. Similar effects have been observed in vivo .
In contrast to examined macromolecules the low molecular solution of GSM (Figure 2A) spread most rapidly in vitreous, justifiable with diffusion-controlled process at initial concentration gradient that provides motive force. Diffusion into more peripheral areas occurred more slowly lasting about 6 h. Visual equilibrium state was approximately achieved at point of 12 h.
Compared to GSM, GSD diffusion zone (Figure 2B) reached peripheral regions of VM at later time points. An equilibrium distribution was not achieved during the period of investigation.
Distribution of dissolved GSP in VM (Figure 2C) and in enucleated porcine eyes (Figure 2D) yielded similar results. The initial injection spots (t=0 h) became less distinct but were present still after 12 h. In either instance the peripheral distribution of the macromolecular GSP was slower than GSM and GSP but in porcine vitreous humor faster than in VM, justifiable with age of animal model and presentational liquefaction. The process of distribution was not completed.
3.2 Distribution studies under simulated posterior vitreous detachment
In clinical routine senior patients are treated with intravitreal injections because of posterior segment diseases, but vitreal degeneration is mostly not considered. In vitro advanced PVD with a liquefied vitreous of 50 %  was simulated by filling the VM with a gelatinous fraction of 50 % as well as Ringer buffer solution. As the injection technique has a distinctive effect on volume and velocity of the drug distribution in liquefied vitreous [4, 11] GSP was either injected through the gelatinous fraction or through the liquid phase. The results are shown in Figure 3.
Following drug injection through PAAgel, hyperintense wavy tracks near the puncture duct remained for at least 6 h and established the small visible volume of distribution in PAAgel (Figure 3A). Applying through liquefied buffer solution (Figure 3B) the mass transport within aqueous area largely overcomes diffusive transport in gelatinous area and induced initial higher distribution in VM. An initial injection track occurred as well (Figure 3, t=0 h). After 12 h a nearly uniform distribution of the macro-molecular contrast agent in VM was detected. The distribution of GSP in gelatinous phase is much slower than in liquid phase, justifiable with the distinctive mesh-work of PAA-gel. Structural components, also present in native vitreous body, impose mixing processes and drug distribution . Additional intravitreal distribution processes, such as hydrostatic pressure, convective flow and aqueous and retinal clearance pathways are incompletely known . At present the VM is limited by the closed circulation processes.
Biorelevant test systems may be helpful to gain a thorough knowledge about distribution processes and intravitreal pharmacokinetics of injections and implants. The investigated combination of VM and MR scanner presents a new approach for the in vitro characterization of drug release and distribution of intravitreal injected dosage forms. The achieved images with high spatial resolution indicated a drug distribution depending on the molecular mass and agerelated liquefaction of the simulated vitreous. Conceivable modified methods are the attachment of animal models to the EyeMoS or varied volume fractions of PAAgel and buffer solution to approximate individual vitreal liquefaction and the situation in vivo. Investigating an MR-suitable EyeMoS provides an insight into the methodical potential of the presented setup.
This project was funded by the Federal Ministry of Education and Research (BMBF) within REMEDIS “Höhere Lebensqualität durch neuartige Mikroimplantate” (FKZ: 03IS2081). Thanks to Malte Bogdahn for his excellent technical assistance.
Conflict of interest: Authors state no conflict of interest. Material and Methods: Informed consent: Informed consent is not applicable. Ethical approval: The research conducted with animal byproducts was compiled in accordance with all the relevant national regulations and institutional policies regarding animal welfare.
Li, S. K., J Lizak, M., & Jeong, E. K. (2008). MRI in ocular drug delivery. NMR in biomedicine, 21(9), 941-956.
Tan, L. E., Orilla, W., Hughes, P. M., Tsai, S., Burke, J. A., & Wilson, CG (2011). Effects of vitreous liquefaction on the intravitreal distribution of sodium fluorescein, fluorescein dextran, and fluorescent microparticles. Investigative ophthalmology & visual science, 52(2), 1111-1118.
Bonfiglio, A., Repetto, R., Siggers, J. H., & Stocchino, A. (2013). Investigation of the motion of a viscous fluid in the vitreous cavity induced by eye rotations and implications for drug delivery. Physics in medicine and biology, 58(6), 1969.
Friedrich, S., Cheng, Y. L., & Saville, B. (1997). Drug distribution in the vitreous humor of the human eye: the effects of intravitreal injection position and volume. Current eye research, 16(7), 663-669.
Loch, C., Bogdahn, M., Stein, S., Nagel, S., Guthoff, R., Weitschies, W., & Seidlitz, A. (2014). Simulation of Drug Distribution in the Vitreous Body After Local Drug Application into Intact Vitreous Body and in Progress of Posterior Vitreous Detachment. Journal of pharmaceutical sciences, 103(2), 517-526
Kleinberg, T. T., Tzekov, R. T., Stein, L., Ravi, N., & Kaushal, S. (2011). Vitreous substitutes: a comprehensive review. Survey of ophthalmology, 56(4), 300-323.
Rosset, A., Spadola, L., & Ratib, O. (2004). OsiriX: an open-source software for navigating in multidimensional DICOM images. Journal of Digital Imaging, 17(3), 205-216.
Zarranz-Ventura, J., Carreńo, E., Johnston, R. L., Mohammed, Q., Ross, A. H., Barker, C., Sallam, A. (2014). Multicenter Study of Intravitreal Dexamethasone Implant in Noninfectious Uveitis: Indications, Outcomes, and Reinjection Frequency. American journal of ophthalmology, 158(6), 1136-1145.
Inoue, M., Takeda, K., Morita, K., Yamada, M., Tanigawara, Y., & Oguchi, Y. (2004). Vitreous concentrations of triamcinolone acetonide in human eyes after intravitreal or subtenon injection. American journal of ophthalmology, 138(6), 1046-1048.
Tolentino, M. (2011). Systemic and ocular safety of intravitreal anti-VEGF therapies for ocular neovascular disease. Survey of ophthalmology, 56(2), 95-113.
Rodrigues, E. B., Grumann Jr, A., Penha, F. M., Shiroma, H., Rossi, E., Meyer, C. H., Farah, M. E. (2011). Effect of needle type and injection technique on pain level and vitreal reflux in intravitreal injection. Journal of Ocular Pharmacology and Therapeutics, 27(2), 197-203.
Wilson CG, Tan LE, Mains J. 2011. Principles of retinal drug delivery from within the vitreous drug product development for the back of the eye. In Kompella UB, Edelhauser HF, Eds. Springer, New York, USA, pp 125–158.