2.1.

Static Light Scattering

Static light scattering (SLS) simply consists of launching light into a system of scatterers and collecting the scattered light. In a transparent system such as pure water, the magnitude of the detected scattered intensity is virtually negligible since water molecules do not scatter light significantly because their molecular size is much smaller than the wavelengths of visible light. However, scattered light intensity increases significantly as the molecular size grows to the same order as the wavelength of the incident light. An increase in the amount of scattered light is an indication of a change in the morphology of the scatterers and therefore can be used to characterize a system qualitatively. This can be used to measure early morphological changes in the cornea, aqueous, lens, vitreous, and possibly retina.

An example of this is shown in Fig. 2 in which an eye of a 2-year-old guinea pig was scanned with a compact fiber optic probe, described later, that collects backscattered photons in vivo. The SLS scan clearly shows the anatomical features of the transparent eye in this animal. Based on the magnitude of the scattered light intensity, the corneal epithelium, stroma, and endothelium, as well as the lens (anterior-posterior) and vitreous are identified. Initially, the intensity rises rapidly at the cornea–air interface owing to a change in the refractive index. It decreases in the stroma and then increases in the endothelium. Local variation in intensity within the stromal structure can be attributed to the scattering of light from uniform collagen fibrils and fibroblasts that are interposed at equal distances.

Figure 2

Static light scattering (SLS) scan showing ocular anatomical features in a guinea pig.

2.2.

Dynamic Light Scattering

The technique of dynamic light scattering provides a quantitative measure of early morphological changes by following temporal changes in and around the random positions of scatterers as they undergo Brownian motion on a time scale of ⩾1 μs. Dynamic information such as diffusion, size, shape, molecular interactions, and flexing motion in polymers can be obtained. This makes DLS a unique tool for probing macromolecular dynamics.⁴ The most attractive features of DLS are that it is noninvasive and quantitative, works effectively for particle sizes ranging from a few nanometers to a few micrometers, requires only a small sample volume, and works reasonably well for polydisperse or multiple-size (up to two to three component) dispersions. DLS is synonymous with quasi-elastic light scattering, photon correlation spectroscopy (PCS), light-beating spectroscopy (homodyne and heterodyne), and intensity fluctuation spectroscopy (IFS). These concepts were borrowed from the techniques used in radio communication.

Figure 3 is an oversimplified depiction of how a DLS measurement is performed in analogy to radio communication. In radio transmission, a microphone converts an audio (voice) signal with an approximate frequency of 256 Hz into electrical pulses. These pulses are then mixed (superimposed) with a carrier wave whose frequency is on the order of 10⁶ Hz. This process is called frequency modulation (FM) or amplitude modulation (AM). The final AM or FM signal is then transmitted via an antenna into the atmosphere. On the receiver end, a radio antenna intercepts the transmitted signal, and if it is tuned to the same frequency as the transmitted signal, it decodes the desired audio signal. At this stage, the opposite of the transmission process takes place and the audio signal is decoupled from its carrier frequency. The desired information is recovered through a speaker diaphragm vibrating at the same audio frequency as the original voice signal.

Figure 3

Dynamic light scattering (DLS) in analogy with radio communication.

In an analogous manner, the laser beam in a DLS measurement can be considered a carrier wave on which the particle (Brownian) motion is superimposed. The particles are often suspended in a fluid. The Brownian motion results from the collisions between the particles and the fluid molecules. On the receiver end (depending on the scattering angle θ) scattered light is collected, decoupled from the laser frequency, and dynamic information about the particles is obtained. This includes information on shape, size, movement, and the interactions among the particles, which can be of vital importance, e.g., in monitoring a cataract as it grows.

3.1.

Compact SLS-DLS Fiber Optic Probes

The fiber optic-based DLS design developments for studies of cataractogenesis are covered by Dhadwal et al.;,⁸ Rovati et al.;,⁹ Ansari et al.;,¹⁰ and the references contained in their papers. The fiber optic SLS-DLS probes, shown in Fig. 4 and diagrammed in Fig. 5, combine the unique attributes of small size and low laser power with high sensitivity [the β values in Eq. (1)]. They were originally developed to conduct fluid physics experiments under the challenging conditions of microgravity on board a space shuttle or space station orbiter. This type of probe is easy to use because it requires neither precise optical alignment nor vibration isolation devices. Currently it has been successfully employed in many ocular experiments to detect early effects of various eye diseases. A low-power (50 to 100 μW) light from a semiconductor laser, interfaced with a monomode optical fiber, is tightly focused on a 20-μm diameter focal point in the tissue of interest via a graded refractive index (GRIN) lens. On the detection side, the scattered light is collected through another GRIN lens and guided onto an avalanche photodiode (APD) detector built into a photon-counting module. APD-processed signals are then passed onto a digital correlator for analysis. The probe provides quantitative measurements of pathologies in the cornea, aqueous, lens, vitreous, and retina. The device is modular in design. If needed, by a suitable choice of optical filters it can be converted into a device for spectral measurements (autofluorescence and Raman spectroscopy) and laser Doppler flowmetry or velocimetry, providing measurements of molecular structure and blood flow in the ocular tissues. The device can be easily mounted onto many conventional ophthalmic instruments to significantly increase their diagnostic power.

Figure 4

Compact fiber optic probes.

Figure 5

Schematic of probe.

3.2.

Sensitivity of DLS Over Established Clinical Technologies in the Early Detection of Cataract

The gold standard in the evaluation of cataract is still slit-lamp biomicroscopy and/or Scheimpflug photography using various clinical grading standards of lens opacity, such as the lens opacity classification system (LOCS), Oxford, Wilmer, or Wisconsin systems. The interpretation of photographs using these grading systems remains subjective. Recent clinical studies using DLS were reported by Thurston et al.;,¹¹ Rovati et al.;,⁹ Dhadwal and Wittpen,¹² and Datiles et al.;,¹³ and the references contained in their publications.

DLS has been found to be far superior in detecting cataract changes very early in the natural history of the disease. But until recently it was not demonstrated just how early a cataract can be detected in a quantitative manner. This was resolved by comparing DLS and Scheimpflug imaging (SI) techniques simultaneously. A detailed description of the SI technique is given elsewhere.¹⁴ The SI technique provides a measure of lens opacity by converting stereoscopic lens images into a turbidity parameter expressed in optical density units (ODU).¹⁴ An ODU value of 0.01 represents the transparency of an ultraclean sample of water. Clinically, a range of values <0.1 ODU is assigned to a normal (noncataractous) lens.

The DLS fiber optic probe described earlier was mounted on an SI system for an objective study of lens opacity.¹⁴ The instrument’s sensitivity was evaluated by slowly inducing a nuclear cold cataract in a calf eye, taking Scheimpflug photographs and DLS data simultaneously (within a few seconds of each other). The γ-crystallins also known as cryo (low-temperature) proteins in a calf lens promote the aggregation of α-crystallins, causing lens opacity as the temperature is lowered. The Scheimpflug images and the DLS data are compared in Fig. 6. The protein particle size distributions consistently shifted from left to right in the direction of increasing sizes at each temperature decline, from 18°C down to 9°C. This can be attributed to the formation of protein aggregates and high molecular weight complexes even though the lens remains transparent. Scheimpflug images, however, did not detect any change until 10°C, when the optical density increased from 0.01 to 0.02 ODU. This value corresponds to N0 on the LOCS-II clinical cataract grading system. The grading ranges from N0 (no cataract) to N4 (mature cataract). The equivalent ODU values are approximately 0.050 (N0), 0.09 (N1), 0.21 (N2), 0.46 (N3), and 0.56 (N4).

Figure 6

Scheimpflug images and DLS size distributions.

One symptom of a developing cataract is glare that is due to increased light scattering. DLS data complement SLS data, as shown in Fig. 7. The scattered light intensity increases as a function of decreasing temperature. This intensity increases slowly at first, with a sharp increase at 14°C, and continues increasing as the temperature continues to drop. In contrast, the ODU from Scheimpflug remains constant until 10°C. At 7°C, ODU measurements increased to 0.06 (error limit: ±0.023 ODU),¹⁴ which corresponds to N0 (normal) on the LOCS-II grading scale. Based on these dynamic and static observations, it can be concluded that DLS is 2 to 3 orders of magnitude more sensitive than the SI imaging system for detecting cataract early.

Figure 7

Comparison of Scheimpflug optical density units (ODU) and DLS scattered light intensity.

3.2.2.

New clinical DLS instrumentation for ophthalmic diagnostics

Until recently, almost all DLS instruments have employed slit-lamp biomicroscopy in one way or another^{8 11} because of its familiarity to ophthalmologists. In our earlier work, we simply mounted the DLS probe discussed earlier on a slit-lamp biomicroscope using a Hruby lens holder. Conceptually, these setups were easy to use and provided ease of alignment, since a test site for DLS measurement was selected by directly observing the lens with the view that “you measure what you see.” However, their ability to sample the same location in the lens repeatedly and reliably is questionable. It is very important to sample from the same location because the lens is not homogeneous and a slight misalignment can give erroneous results, especially in repeat patient visits during longitudinal studies. New instruments have been developed to circumvent these problems by combining DLS with other ophthalmic instruments employing such technologies as autofluorescence and keratoscopy. The advantage of this is that not only can the sampling issue be resolved but DLS data can also be compared with other complementary ocular data, e.g., oxidative stress, as shown in the review article on autofluorescence by Rovati and Docchio in this issue.

A corneal analyzer (Optikon 2000 Keratron, Italy) utilizes an infrared (IR) detector that pinpoints the exact location of the corneal vertex. The front-end eyepiece is equipped with two fiber optic position sensors on the extreme (equatorial) ends. An IR beam passes through one fiber and is collected by another fiber at the opposite end. When this beam is intercepted by the cornea (see Fig. 8 and Fig. 9), the instrument is aligned with the apex of the cornea. This instrument was modified by placing a DLS fiber optic imaging probe inside its cone as shown in Fig. 8 and Fig. 9. The DLS portion of the instrument takes measurements from the cornea to the retina by moving the DLS probe inside the cone using computer-controlled linear microactuators. Since the IR beam does not pass through any other ocular tissue, the actual laser light used during the 2 to 5-s DLS measurement is never used to align the instrument, as was the case in previous slit-lamp-based DLS instruments. This arrangement makes it much safer for human use. During the DLS measurement, the subject is asked to fixate on a light source housed inside the cone and probe assembly.

Figure 8

Front and cutaway view of a modified corneal analyzer (Keratron) with a DLS probe.

Figure 9

Schematic diagram of the optical system.

Using the CCD camera located in the center of the cone (Fig. 9), an operator uses a joystick to center the cone with the optical axis of the eye, in order to view the active image of the patient’s eye on the instrument’s built-in CRT screen, as shown in Fig. 10. When the foot switch is depressed, the IR detection system automatically sends a signal to the computer, instantaneously grabbing the image when the desired position of the cone has been achieved. The operator simultaneously depresses the “remote DLS” button, which in turn activates the laser and starts the digital autocorrelator for the DLS measurement. The patient sees a diffused speckle light pattern during the 2 to 5-s DLS measurement. The laser and the data acquisition program can be preset to run for either a 2-s (for the cornea) or a 5-s (for the lens) measurement time. The operation sequence is shown in Fig. 11.

Figure 10

Instrument in clinical operation.

Figure 11

Block diagram of data acquisition and analysis.

This new instrument has demonstrated its reproducibility and ability to go back to the same location in the lens during repeat patient visits.¹³ Since that study, about 50 patients have been studied at the National Eye Institute in Bethesda, Maryland. None reported any discomfort during or after the DLS measurements. The data will be useful in planning a longitudinal study of anticataract drug(s) screening at the NEI.

4.1.

Cornea

Corneal tissue is avascular and composed primarily of collagen. The human cornea is about 500 μm thick at the apex and about 700 μm at the periphery or limbus.¹⁵ Transparency is the most important corneal property in maintaining good-quality vision. A slight loss of transparency can cause problems such as haze and glare. Histologically the cornea is structured as a stack of layers divided into five distinct regions. These include the epithelial cell layer, Bowman’s layer, stromal layer, Descemet’s membrane, and the endothelial cell layer. The epithelial layer (about 50 μm thick) consists of six to eight layers of cells of which two to three layers are superficial flat cells, followed by a squamous, irregularly shaped wing (about two layers), and a single layer of cylindrical basal cells. Bowman’s membrane (about 10 μm thick) is a dense, irregular meshwork of interwoven collagen fibrils of type I and type VII without a cellular structure (no fibroblasts). The stromal layer comprises 90 of the thickness and volume of the cornea, consisting of uniform-diameter collagen fibrils (about 200 in number) with regular spacing. Fibroblasts are interposed in these collagen layers. Descemet’s membrane is a tough, glassy layer that can vary in thickness from about 5 to 15 μm as a function of age. The endothelium consists of a single layer of metabolically active, irregular polygon-shaped cells about 20 μm in diameter.¹⁵ The tear film (about 7 μm thick) on the surface of the epithelium prevents the cornea from dehydrating.

The use of SLS and DLS in the cornea offers opportunities and challenges. The opportunity lies in its transparency, since the normal cornea does not exhibit multiple scattering of light. However, at first glance, the different layers and structures within the cornea seem very challenging when one is interpreting DLS data. Furthermore, corneal tissue is far from the ideal case of a dispersion containing colloidal particles. One should therefore refrain from interpreting corneal DLS data in terms of particle size because of many unknown quantities (local viscosity, refractive index, etc.). However, it is reasonable to report the decay constant Γ (s⁻¹) from Eq. (1), which represents the time relaxation values of a scatterer as it executes limited Brownian excursions about fixed average positions in the corneal tissue.

4.1.1.

Brownian spectrum of the cornea

As described in Sec. 2.2.1, the construction of a TCF [Eq. (1)] in a DLS experiment relies heavily on measuring the fluctuations of scattered light intensity in a time domain. To demonstrate this point, an experiment was performed on a model system of polystyrene microspheres in the size range of 70 nm to 1 μm suspended in water, and the time fluctuations were compared with the scattered light intensity as a function of particle size. The spectrum of the microspheres is illustrated in Fig. 12(a), for 70-nm, 250-nm, and 1.05-μm particles. As expected, the time fluctuations in this freely diffusing system of particles gradually slow down as the particle size increases. The time decay constant (Γ) shown in Fig. 12(b), which was obtained directly from the TCF, further illustrates this point. It shows that small particles move very fast while larger particles move very slowly and in a random fashion around their fixed average positions. Thus, Γ becomes a powerful parameter when one is characterizing a system of freely diffusing and bound particles without prior knowledge about the system (e.g., viscosity, refractive index, size).

Figure 12

(a) Intensity fluctuations and (b) Brownian motion.

Figure 13 compares the intensity fluctuations within the layers of a normal bovine cornea in an intact eye, in vitro. The components in the anterior stroma fluctuate slowly, while in the middle and posterior sections the fluctuations are relatively faster. The endothelial cells fluctuate much more slowly. When the laser light enters the anterior chamber, the fluctuations almost disappear. Thus we can infer that either the clear aqueous humor in the bovine eye contains hardly any particles or proteins of a size to scatter light or the power level of the probing laser light of (a few microwatts) is too low, resulting in the intensity of scattered light being too low to be detected. In either case, the power level of the scattered light is too low to construct a TCF or a spectrum. However, as described later, protein concentration in the human aqueous can increase during uveitis.

Figure 13

Brownian spectrum of bovine cornea.

The polygon-shaped cells in the endothelium layer are rigidly bound and act like a reflecting mirror. As a result, the entire structure exhibits very slow Brownian motion. The faster fluctuations in the middle and posterior stroma can be attributed to the presence of protein particles and an increased amount of water. Indeed, the water content increases from the anterior to the posterior cornea, as shown by Castoro et al.;,¹⁶ using differential scanning calorimetry. This may be why the fluctuations in the anterior stroma are relatively slower. In terms of post-LASIK complications, since the epithelium layer is preserved in a flap, the stromal region can be of prime interest for examination with DLS.

4.1.3.

Post-LASIK results

Two sets of DLS data are presented after LASIK surgery was attempted on two bovine eyes in vitro. Ronald Krueger, M.D. of the Cleveland Clinic performed these procedures. They are referred to as eye 1 (photorefractive radial keratectomy, PRK) and eye 2 (lamellar keratectomy, LK) and are shown in Fig. 14 and Fig. 15, respectively. The DLS measurements on the untreated cornea were repeated five times to get an estimate of variation (see error bars) in Γ values. The bovine corneas were found to be much more “rigid” than the human corneas, so an attempt to make a flap using a lamellar keratectomy of about 3 nm was not successful. In eye 1 the epithelium was scraped and the PRK procedure (about 8 mm in size, 5.5 diopters) was performed. In eye 2 the epithelium was scraped and two lamellar karatectomies about 5 mm (no laser ablation) with a residual thickness of about 720 μm were performed. Since the corneas are reduced in size after the surgery, the DLS data for treated corneas begin at a distance of 0.2 and 0.25 mm, respectively. The two surgically treated corneas show higher Γ values than their untreated counterparts. The PRK and LK results differ significantly. The PRK shows less increase in Γ values than the LK. This is perhaps due to the loss of water during the ablation process in the case of PRK. Since no laser ablation was performed in LK, the Γ values remain higher. It is of interest to note that for the most part, Γ values revert back close to untreated levels 1 day after surgery.

Figure 14

Eye 1, photorefractive radial keratectomy (PRK), about 8 mm.

Figure 15

Eye 2, lamellar keratectomy (LK), about 5 mm.

The eyes were kept in a jar of saline solution and refrigerated overnight after the surgeries. The LK procedure showed faster “healing” (return to baseline levels) than the PRK. After corneal injury, as in surgery, the area where the epithelium has been removed is reepithelialized by the sliding movement of the expanded remaining epithelium. Two factors affecting this healing process are the size of the denuded area and the smoothness of the denuded surface, as well as the presence or absence of cellular and other debris on the surface. The larger the denuded surface, the longer it takes to reepithelialize. Since the LK surface was only 5 mm, compared with the PRK (8 mm), it healed faster. The laser-ablated (PRK) surface will still have significant debris on the surface, whereas there will be almost no debris on the lamellar keratectomized surface (LK); this difference might explain why there seems to be more healing in the LK-treated cornea. These differences can be seen by the more rapid return to normal, baseline levels of the Γ values in the LK-treated cornea 1 day after the injury, compared with the PRK-treated cornea. However, more studies need to be performed to confirm these observations, such as using live animal models to follow the healing process using vital stains after LK scraping of the epithelium. The treatments discussed here also changed the corneal refractive power.

The results of these experiments suggested the potential of SLS and DLS in measuring changes in pre- and postrefractive surgeries. However, the actual feasibility of using the DLS technique diagnosis and treatment for clinical in early evaluation of corneal complications after LASIK surgeries and other corneal abnormalities has yet to be proven. At the time of this writing, this validation work is being conducted at NEI/NIH.

4.2.

Aqueous, Uveitis, and Glaucoma

Aqueous humor can be considered an ultrafiltrate of blood since it contains most of the molecules found in serum at concentrations that are reflective of serum levels. For example, certain metabolites (glucose) and proteins (human serum albumin, HSA) are always present in the aqueous. Since glucose molecules are much smaller than the wavelength of light, they do not contribute significantly to scattered light. Because HSA particles can scatter light appreciably, since they are spherical in shape and their size (about a few nanometers) is comparable to the wavelength of light, they are frequently used in DLS experiments. Uveitis is an inflammatory disease of the eye. When the anterior chamber is involved, uveitis can produce high numbers of protein particles in the aqueous humor. This increase in protein concentration gives rise to increased scattered light intensity or “flare” in slit-lamp biomicroscopy. Clinically, flare and therefore the severity of uveitis, can be easily detected and quantified by SLS and DLS. Similarly, cholesterol levels in the body can also be evaluated through the aqueous, as suggested by Bursell.⁶ This important application has not been pursued thoroughly, either in laboratory or clinical settings.

Glaucoma is a disease in which the intraocular pressure (IOP) is often increased. If IOP increases are not treated, they can lead to degeneration of the optic nerve and eventually loss of vision. Recently, DLS was used to study one specific type of glaucoma known as pigmentary dispersion glaucoma (PDG). The posterior layer of the iris epithelium contains many melanin granules. These granules are released into the aqueous as a consequence of natural epithelial cell death, infection, or trauma. Excess melanin release into the aqueous occurs in patients with PDG. The melanin granules can be trapped in the trabecular meshwork, over time blocking aqueous outflow and causing an increase in IOP. Pollonini et al.;²¹ used DLS to detect and quantify aqueous melanin granules in a clinical study of normal and PDG volunteer patients. The results indicated the presence in the aqueous of small particles (1 to 10 nm) in normal patients and large particles (up to 1 μm) in PDG patients and therefore showed the potential of DLS in monitoring this type of glaucoma noninvasively.

4.3.

Lens and Cataract

Cataract (lens opacification), its clinical evaluation, and its classification is discussed in detail elsewhere by Datiles and Ansari.²² Snellen acuity charts, contrast sensitivity tests, visual function questionnaires, examinations with direct ophthalmoscopes, slit-lamp biomicroscopy, Scheimpflug imaging, and retroillumination imaging techniques are routinely employed in cataract evaluation. However, these methods fail to detect and grade cataracts reliably and quantitatively at their early stages of formation. The validation of DLS techniques in animal models is a logical first step before human trials.

4.3.1.

Cataractogenesis in small animal models

Animal models are usually helpful in the validation of new instruments and testing of therapies prior to clinical use. Until recently, it was difficult to use SLS and DLS in small animals (e.g., mice) mainly owing to instrument limitations (e.g., physical size, power requirements, and alignment problems). Compact DLS probes developed at the National Aeronautics and Space Administration (NASA) and shown in Fig. 4 are proving to be very useful in this area of research. A generic experimental setup is shown in Fig. 16. Cataractogenesis is monitored in vivo by following TCF profiles [Eq. (1)] on different time lines. As an example, Auto-Correlation Profiles for Philly mice are shown in Fig. 17(a). This animal develops cataract spontaneously between days 26 and 33 after birth. The data include a 45-day-old normal mouse of the control FVB/N strain, which does not develop a cataract, and two Philly mice roughly 26 to 29 days old.

Figure 16

SLS-DLS scanning setup for small animals.

Figure 17

In vivo cataract measurements in Philly mice. (a) Autocorrelation profiles. (b) Normal particle size distribution, control mouse. (c) Particle size distribution for a mouse with trace cataract. (d) Particle size distribution for a mouse with mild cataract.

Examinations of these mice with a slit-lamp biomicroscope identified one normal (transparent) eye and two other eyes with trace cataracts. These examinations were conducted soon (within few minutes) after the DLS measurements were completed. Each measurement took 5 s at a laser power of 50 μW. The changing TCF slope is an indication of cataractogenesis as the lens crystallins aggregate to form high-molecular-weight clumps and complexes. The DLS autocorrelation data are converted into particle size distributions [Figs. 17(b) to 17(d)] using an exponential sampling program. Although conversion of the DLS data into particle size distributions requires certain assumptions regarding the viscosity of the lens fluid, these size values do indicate an increasing trend as the cataract progresses. These measurements suggest that a developing cataract can be monitored quantitatively with reasonable reliability, reproducibility (5 to 10), and accuracy.

4.3.2.

Off-axis early cataract mapping

Light falling on the cornea is guided by the pupil along the optical axis of the eye and is focused on the fovea by the lens. Any lens abnormality along this optical path, whether capsular, nuclear, or cortical, can potentially deteriorate the sharp focus of the image on the retina and thus vision. Therefore every DLS study of cataract has been carried out along the optical axis of the eye. Until recently, no DLS data were available for the peripheral regions of the lens. There is a debate among cataract researchers regarding the causes of subcapsular cataracts, nuclear cataracts, and cataracts occurring in the peripheral regions of the lens. Thus there is a need to gather data on the early onset of opacities simultaneously in different regions of the lens.

Ansari et al.;²³ used DLS in vivo to map the lenses of HIV-1 protease transgenic mice to study early onset of cataractogenesis. Topographical colored contour maps (cataractograms) were obtained noninvasively by automated scanning of the eye with a DLS probe in three dimensions (x, y, and z directions). Various colors in the map represent varying crystallin size and provide quantitative detection and monitoring of opacities in the entire lens. The portion of the lens mapped is, however, limited by the degree of pupil dilation (∼1 mm) for the mouse eye. Four representative cataractograms are presented in Fig. 18 for transgenic mice ranging in age from 17 to 23 days. These images in false color (black to white) provide topographical contour maps of the lens that show the degree of severity of a progressing cataract. The cataractograms show a tiny amount of opacity developing on day 17, a small opacity occurring in the posterior region on day 22, and a dramatic increase on day 23, becoming large by spreading into nuclear and anterior segments of the lens. It is also interesting to note the change in the average scattered intensity as a function of increasing particle size. The average size changed from a minimum of 30 nm to 662 nm and intensity changed from 100 kilocounts/s to 1500 kilocounts/s. The average particle size increased 20-fold and the scattered light intensity increases 15-fold from day 17 to day 23. However, the normal strain does not develop a cataract, as shown for an 18-day-old normal mouse.

Figure 18

Cataractograms (cataract mapping).

The behavior we see here is very different from the formation of large protein complexes during the progression of a nuclear cataract, in which the particle size or intensity is expected to increase only in the central portion of the lens. In a study of HIV-1 protease transgenic mice, Tumminia et al.;²⁴ found that cataracts appear between 24 and 26 days after birth. Bettelheim et al.;²⁵ viewed hydration as a main cause of cataract formation in these animals, based on differential scanning calorimetry and thermogravimetric analysis. It is therefore clear that different mechanisms can be responsible for the formation of cataracts in humans and other animals. In summary, DLS was able to identify a progressing cataract under these widely different and distinct mechanisms of formation.

4.3.3.

Human congenital cataract

As mentioned earlier, the present cataract grading systems are only applicable to age-related or senile cataracts and are thus not useful for congenital cataracts. Congenital or infantile cataract is a lens opacity that is present at birth. These types of cataracts are classified by their morphology, presumed etiology, the presence of certain metabolic disorders, or associated ocular or systemic findings.²⁶ The ability of DLS to study such opacities quantitatively in humans is discussed here. Figure 19(a) shows a slit-lamp photograph of the right eye of a 37-year-old male with a mild congenital nuclear cataract. It is located in the embryonic nucleus slightly off of the optical axis. The left eye of this volunteer patient was found to be transparent. The DLS measurements were taken about 6 mm from the cornea into the lens, both on and off the ocular optical axis. The particle size distributions plotted in Fig. 19(b) clearly show the presence of smaller protein particles in the clear section of the nucleus and aggregates of much larger particles near the visual opacity. The size distribution for the left eye is comparable to that of the right eye in the nucleus and is consistent with age-related changes for a similar age group.

Figure 19

Human congenital cataract. (a) Slit-lamp photograph. (b) Particle size distributions obtained by DLS. (See text for explanation.) (Slit-lamp photograph courtesy of M. B. Datiles, National Eye Institute/National Institutes of Health.)

4.3.4.

Developing pharmacotherapy using SLS and DLS

In the absence of a nonsurgical treatment, 50 of all blindness is due to cataracts. At present 34 million Americans over the age of 65 have cataracts. Foster predicts this figure to rise to 70 million by year 2030.²⁷ According to Kupfer,²⁸ “A delay in cataract formation of about 10 years would reduce the prevalence of visually disabling cataracts by about 45.” DLS can help achieve this goal by detecting cataracts at very early stages and by screening potential anticataract drugs for their efficacy.

Anticataract drug screening. The results of both preclinical and clinical tests of potential therapeutic agents indicate the need for more sensitive measures of protein aggregation and opacification in vivo. Studies in humans predict that DLS will be useful in testing anticataract drugs to inhibit or reverse the progression of cataract formation during longitudinal clinical trials.^{11 12 13 29 30} Pantethine is a metabolically active, stable disulfide form of pantetheine and a derivative of pantothenic acid (vitamin B ₅) . It can be used as an anticataract drug, but the clinical trial of pantethine was inconclusive with respect to effectiveness.^{29 31}

Recently Ansari et al.;³² using SLS and DLS demonstrated the efficacy of pantethine in very early stages (a few hours to a few days) of selenite-induced lens damage, well before the formation of a mature cataract in the selenite model.³³ In this study, SLS and DLS measurements were made on 33 animals (rats) aged 12 to 14 days at the time of selenite or pantethine injection. Pantethine treatment resulted in a substantial decrease in the dimensions of scattering centers in lens in vivo during the early stages of opacification prior to their detection with conventional ophthalmic instruments. An obvious opacity was not observed using a slit-lamp biomicroscope. Typical representative size distribution data are presented in Fig. 20 for a control, Se-treated (12 and 60 h post-treatment), and Se-pantethine-treated animal (42 h post-treatment) at a distance of about 2 mm from the anterior surface of the lens. The size distribution of scatterers in the control animal ranged from 60 to 80 nm in diameter. In the Se-treated animals, the TCF shows bimodal behavior with aggregates ranging in size from 400 to 800 nm and 3000 to 10,000 nm in diameter. The Se-pantethine-treated animal shows an aggregate size averaging around 200 nm. The experimental results clearly indicated that SLS and DLS were able to discern subtle molecular changes very early in cataract formation.³⁴ The results suggested that pantethine inhibits the initial aggregation process.

Figure 20

Cataract treatment in rats. Analysis of particle size distribution with exponential sampling.

In conclusion, the results are encouraging. The future outlook for finding a medical cure for cataract seems optimistic when this effort is combined with advances in detection technology, understanding of fundamental processes, and the efforts in designing new anticataract drug formulations. This work is picking up momentum at Zigler’s laboratory at the NEI/NIH.^{35 36 37} In a recent pilot study, Congdon’s group at Johns Hopkins University has shown that isoflavones can prevent or delay X-ray-associated cataract.³⁸ Also, Japanese investigators have demonstrated a delay in cataract in rats receiving miso (a soy-containing product) compared with control animals.³⁹

4.4.

Vitreous and Diabetic Vitreopathy

The vitreous is the least understood part of the eye since it remains transparent throughout life. Sebag discusses the role of vitreous in pathology and the technologies to image this transparent tissue in his review article in this issue. Ansari et al.;⁴⁰ and Rovati et al.;,⁴¹ were the first to characterize vitreous structure with DLS and correlate the findings with diabetic retinopathy. In a DLS measurement, the vitreous body exhibits a two-exponential behavior consistent with its gel-like properties. Equation (1) can be rewritten in terms of a fast- and a slow-decaying component:

Figure 21

Time relaxation of human vitreous.

Figure 22

Liquid bovine vitreous.

5.1.

Diabetes

DLS can be used to detect early onset of diabetes through the transparent ocular lens. Chenault et al.;⁴³ have indicated that cataracts are 1.6 times more common in diabetic patients than non-diabetics. The U.S. Food and Drug Administration (FDA) has a unique animal model for studying type II diabetes. Psammomys obesus (sand rat) is a wild rodent found in the desert areas of the Middle East and Africa. It develops mild to moderate obesity, hyperglycemia, and the complications of diabetes, such as cataracts and vision loss, when it consumes a high-calorie diet. In a recent study,⁴³ blood glucose levels, insulin, and glycohemoglobin values in this animal were correlated with histopathology, traditional ophthalmology assessments, and DLS measurements with a view toward developing a noninvasive means to detect early stages of eye damage that is due to diabetes mellitus. The control animals demonstrated no significant changes in particle size distributions. The study animals exhibited significant increases in the sizes of (crystallin) scattering centers and substantial shifts in the size distributions. No discernible difference between the control and study animals was observed by ophthalmic examination or by histology. The DLS results revealed subtle changes in the lens of the diabetic sand rats after only 2 months on the diabetogenic diet [Fig. 23(a)]. Conventional ophthalmic instrumentation did not detect these changes [Fig. 23(b)]. At the time of this writing, experiments using sand rats and SLS and DLS to screen some conventional and nonconventional drugs to control diabetes (hyperglycemia) and lens damage are under way at the FDA.

Figure 23

Cataract damage that was due to diabetes in sand rats. (a) Particle size distribution as determined by DLS. (b) Slit-lamp photographs of eye. (Photos courtesy of Michelle Chenault, U.S. Food and Drug Administration.)

5.2.

Alzheimer’s Disease

Like the ocular diseases described earlier, the incidence of Alzheimer’s disease (AD) is expected to take on epidemic proportions in the next two to three decades. Can it be detected through the eye? This sounds like a radical idea because AD is a disease of the brain and not the eye. Frederikse et al.;⁴⁴ at NEI/NIH were the first to demonstrate that Alzheimer’s biology occurs in the lens. AD sufferers lose cognitive abilities and thus experience serious deterioration in the quality of life. In simplistic terms, amyloid proteins forming plaque on the brain tissue lead to development of AD. Currently, these amyloid deposits can only be studied under a microscope by looking at the brain tissue during autopsy. New in vivo experimental methods under investigation include positron emission tomography (PET) and magnetic resonance imaging (MRI), but their spatial and contrast resolution for very early detection of AD is questionable. Nevertheless these techniques are being used to follow treatment regimens in AD patients.

Recently, Frederikse provided evidence of amyloid protein structure in the lens and its association with cataractogenesis.⁴⁵ It was suggested by Goldstein et al.;⁴⁶ that amyloid proteins may promote the aggregation of ocular proteins, linking the formation of supranuclear cataract in human lenses to AD. If this is true, then DLS can be used in the eye to detect early symptoms of AD. A proof-of-concept experiment was conducted in vivo by measuring protein aggregation in the lenses of transgenic mice. The results shown in Fig. 24 indicate enhanced aggregation in study animals as opposed to controls. The biochemical analysis conducted at autopsy and presented elsewhere supported the conclusions based on DLS.⁴⁶

Figure 24

Protein aggregation in the lens of transgenic mice.

It seems that the ocular signs of AD are not confined to the lens. In a recent study by Janciauskiene et al.;,⁴⁷ AD was linked to glaucoma and pseudoexfoliation syndrome by identifying Alzheimer’s amyloids in the aqueous humor using electrophoresis followed by Western blotting. This opens up an opportunity for DLS to be used noninvasively in detailed studies of AD treatment using antiinflammatories, antioxidants, and hormone replacement therapies.

7.1.

Subject Variability

The lens and vitreous, being transparent, do not have any visible markers. Therefore, in longitudinal studies, reproducibility in repeat patient visits remains a big concern. DLS instrumentation must have the ability to precisely control and sample the same location inside the tissue of interest. The eye movements must be controlled and the subject should remain well fixated during the DLS measurement cycle. In animal experiments, this is not much of a concern because animals can be sedated. However, they should be closely watched since some animals roll their eyes while sedated.

7.2.

Safety

In DLS ophthalmic applications, light from a laser source in the visible-infrared spectrum is used. Therefore the major safety concern is the amount of light exposure on the retina. To minimize the risks, the power levels must be low and the exposure time must be short to satisfy the safety requirements set by the American National Safety Institute. Current ophthalmic DLS instruments typically use a laser power of roughly 50 to 100 μW, with an exposure time of 5 to 10 s. This makes it several orders of magnitude lower and therefore safer for use in animals and humans. However, every DLS instrument has its own unique optical arrangement for launching and receiving light signals, alignment procedures, and maintaining coherence conditions inside the scattering volume. Therefore extra caution and care must be exercised in calculations for maximum permissible exposure (MPE) limits set by ANSI.

7.3.

Corneal Dryness and Body Temperature

Drying of the cornea and the appearance of cold cataract produce artifacts, especially in animal measurements. Corneas should be constantly irrigated with saline solutions between the DLS measurements to avoid dehydration. In general, the body temperature of animals under anesthesia drops. This can cause cold-induced cataract, especially in mice and rats. This can be avoided by using a heating pad and keeping the body temperature constant around 37°C during the DLS experiments.

7.4.

Evaluation of Visible Abnormalities

It is tempting to compare the DLS data, especially in clinical practice, with obvious ocular abnormalities. The techniques of SLS and DLS are only suitable for studies of very early ocular abnormalities. They are not suitable for studying visible or mature lens or corneal opacities. Such measurements will introduce artifacts that are due to multiple scattering of light, thus showing diffusion narrowing. The DLS data, especially in transparent lenses, when converted into particle size distributions, show quite nicely how a cataract is progressing. DLS can be used effectively in the determination of accurate particle sizes in water-based dispersions from transparent to extremely turbid (about 7 orders of magnitude higher than the turbidity of water) samples ranging in size from few nanometers to almost a micron. However, such analysis should be avoided in a cataractous eye because of unknown factors such as the changing viscosity. It is therefore acceptable to exploit the good dynamic range of DLS to follow a systematic trend as a marker for monitoring early ocular pathology, but due caution must be exercised in interpretating data when dealing with mild to advanced pathologies.