The mechanism by which vitreous degeneration happens remains uncertain, although a photochemical reaction has been proposed to be a potential underlying cause (Ueno et al. 1987; Akiba et al. 1994). In the studies, the authors demonstrated that visible light excites riboflavin to generate radicals and oxygen species including the superoxide anion, hydrogen peroxide, hydroxyl radicals and singlet anions. These components are known to account for the degradation of hyaluronan and cross-linking of calf collagen in vitro (Akiba et al. 1994; Kakeshi et al. 1994) and in vivo (Ueno et al. 1987). Riboflavin is a photosensitiser that exists naturally in the vitreous environment; the authors proposed that light exposure could be the mechanism underlying the occurrence of age-related vitreous degeneration. Another hypothesis is based on the work by Akiba et al. (1995), who showed that whole serum and a combination of serum protein (transglutaminase) and fibronectin promotes collagen cross-linking leading to vitreous gel contraction. The leakage of plasma proteins from the intravascular space into the vitreous body is possible due to vascular incompetence associated with ageing retinal and ciliary body vasculature. As a result, the concentration of soluble proteins present in the vitreous increases from approximately 0.5–0.6 mg/mL at ages 13–50 to 0.7–0.9 mg/mL at ages 50–80 and 1.0 mg/mL above 80 years (Sebag 1989). In addition, age-related increase in proteolytic activities within the vitreous may also be a contributing factor to vitreous liquefaction (Thomas et al. 2000). The concentration of plasmin, a proteolytic enzyme in the vitreous, increases with age, possibly caused by tissue degeneration such as the retina. In the vitreous, plasmin may combine with membrane type matrix metalloproteinase-1 (MMP-1) to activate progelatinase-A (proMMP-2), which has been documented to have the capability to cleave off hybrid type of V/XI collagen and liquefying the vitreous gel in vitro (Brown et al. 1996).

6.7.2  Physical Changes Involved in the Ageing Vitreous

Balaz and Denlinger established the progression of human vitreous liquefaction in post-mortem biopsy tests performed on 610 human eyes aged between 5 and 90 years (Balazs and Denlinger 1982). The volume of vitreous gel and liquid phase were measured and related to the age of the eye. It was found that the vitreous of young human adults of around 20 years of age was 80% gel phase, which decreased to almost 50% beyond 60 years. The decrease in gel volume was accompanied by a parallel increase in liquid volume as illustrated in Fig. 6.8.

Sebag and Balaz illustrated the changes on maturation using dark-field microscopy illuminated with a slit lamp (Sebag 1987, 2005). Human vitreous from donors, aged 53–88 years, were dissected from the sclera, choroid and retina with the anterior segment remained attached. The non-fixed vitreous was mounted in a transparent chamber containing isotonic saline and sucrose (3.5 g/L) and trans-illuminated. Bundles of parallel and thick fibres coursing along the anterior–posterior direction as well as areas of liquid pockets were seen in the vitreous of a middle-aged man (Fig. 6.9).

Fig. 6.8  Age-related changes in the volume of the vitreous gel and liquid phase. The sustained onset of liquefaction throughout life is evident (adapted from Balazs and Denlinger 1982)

Fig. 6.9  Vitreous structural changes with age. (a) Vitreous humour of a 6 year-old child (4×) (adapted from Sebag 1987, with permission); (b) vitreous humour of a 59-year-old adult (8.3×); (c) vitreous humour of a 88-year-old adult (2.7×). The lower part of the image indicates the position of the lens (adapted from Sebag 2005, republished with permission of the American Ophthalmological Society)

These structures become more prominent in an older person (80–90 years) where fibres were no longer parallel and linear in shape but rather tortuous and broken. Additionally, enlarged liquid pockets in areas devoid of collagen fibres were seen at the central and peripheral areas of the vitreous. This suggests that disruption of the fibrous structure and advanced liquefaction leads to eventual collapse of the whole vitreous; observations which can be explained by changes to the organisation of the vitreous components. Chondroitin sulphate, hyaluronan and opticin, which previously filled the space in between fibrils, are dissociated from the collagen fibrils leading to its lateral aggregation into bundles of fibres. As a result, areas devoid of collagen

fibrils are filled with liquid vitreous containing depolymerised hyaluronan and other soluble substances (Bishop 2000). As opposed to a juvenile vitreous, the elderly vitreous is more fibrous in appearance and smaller as a result of syneresis (Fig. 6.9).

6.7.2.1  Pre-Clinical Model of Ageing Vitreous

Despite the wide recognition of vitreous degeneration with age, pre-clinical drug development for retinal therapeutics is generally conducted in young laboratory animals with an intact vitreous structure. Studies have demonstrated that the percentage vitreous gel content of the laboratory Dutch-belted rabbits (3 months to 2 years old) was ~20% (Tan et al. 2011), a measurement similar to that established by Balaz and Denlinger for young human adults (Balazs and Denlinger 1982). This suggests that laboratory rabbit is a relevant model for younger populations but may not be representative of the elderly eye.

Based on the fact that an elderly model will be useful for ocular disposition studies, our group has established a rabbit model with partial vitreous liquefaction using ovine testicular hyaluronidase. The generated degree of vitreous liquefaction was representative to that seen in the elderly of age around 60 years. The enzyme-induced liquefaction approach was demonstrated to be reproducible without gross ocular tissue changes observed using fundus examination. The model was successfully utilised in assessing intravitreal drug disposition of different molecular weight fluorescent compounds of which results will be discussed in next section (Sect. 6.7.2.2).

6.7.2.2  Effects of Vitreous Liquefaction on Intravitreal Drug Delivery

The effects of vitreous liquefaction have been evaluated on the distribution kinetics of sodium fluorescein, fluorescein dextran (FD) 150 kDa and 1 mm fluorescent particles (Tan et al. 2011). In the study, it was found that sodium fluorescein (MW ~ 376 Da) and FD 150 kDa were distributed and cleared faster from the partially liquefied vitreous as compared to normal vitreous as illustrated in Figs. 6.10 and 6.11, respectively. The faster rate of clearance in the liquefied vitreous suggested that the capability of the vitreous in retaining small and large molecules has significantly reduced and injected substances were expected to have a shorter intravitreal half-life. Nevertheless, ocular fluorophotometry data revealed similar gradient pattern of fluorescent probes along the optical axis in both normal and liquefied vitreous. This shows that although the rate of clearance has accelerated, the elimination pathway by which molecules were cleared was not affected by the vitreous state. In case of microparticles, distribution was found to be more dispersed in the liquefied eye and a faster rate of particle sedimentation was observed as shown in Fig. 6.12. Findings based on these data led us to conclude that vitreous diffusivity and convective forces were enhanced in the partially liquefied vitreous leading to a faster rate of drug clearance.

Clinically, the increased vitreous diffusivity in the liquefied vitreous has been illustrated by Moldow et al. using a fluorescein profile of a 52-year-old patient.

Fig. 6.10  HRA images showing the distribution of sodium fluorescein at 2 and 5 h following intravitreal injection. After 5 h, the fluorescent mass remained well retained in the normal vitreous but appeared to be more diffuse in the liquefied vitreous, suggesting increased vitreous diffusivity (adapted from Tan et al. 2011, with permission)

Fig. 6.11  HRA wide-angle images of intravitreally injected fluorescein dextran 150 kDa in the normal and liquefied vitreous models from 1 h to day 30 after injection. The amount of FD 150 kDa remained in the vitreous was lower in the liquefied vitreous as compared to normal from day 6 onwards, suggesting that FD 150 kDa has a shorter intravitreal half-life in a more liquefied vitreous environment (adapted from Tan et al. 2011, with permission)

The diagnosis was retinitis pigmentosa with vitreous liquefaction or detachment (Moldow et al. 1998). The lack of diffusional gradient across the vitreous cavity observed 30 min after injection into a superficial arm vein could partly be attributed to lower vitreous diffusivity. Additionally, Spielberg and Leys have reported that older patients (mean age: 68.5 years) treated with intravitreal bevacizumab for myopic