Scleral Hydraulic Conductivity and Macromolecular Diffusion in Patients with Uveal Effusion Syndrome

Abstract

purpose. To determine whether uveal effusion syndrome (UES) is caused by altered scleral permeability to water and large molecules.

methods. Transscleral water movement was measured using surgically removed sclera clamped in a modified Üssing chamber and connected to a water column set at intraocular pressure. Sclera was also clamped between two hemichambers, and transscleral diffusion of FITC-dextrans (4.4–77 kDa) was measured with a spectrophotometer. Clinical data were prospectively collected using postal questionnaires.

results. Ten patients (mean age, 63 years; mean spherical equivalent, +4.7 D) had a median preoperative visual acuity of 0.20 that improved to 0.33 after surgery. Nine eyes showed visual improvement, three worsened, and two were unchanged. Histology showed disorganization of collagen fibrils, with amorphous deposits expanding the interfibrillary spaces. The mean thickness (±1 SD) of the excised scleral specimens was 585 ± 309 μm, and the mean specific hydraulic conductivity was 23.9 ± 27.5 × 10⁻¹⁴ cm², compared with 5.8 ± 3.9 × 10⁻¹⁴ cm² in age-matched control specimens (P = 0.068). Three specimens had hydraulic conductivity above the 95% CI of the controls. Control eyes showed a significant reduction in diffusion coefficient (D) with age. Eyes had a mean D of 5.69 ± 5.35 × 10⁻⁸ cm² · s⁻¹, similar to control eyes (6.14 ± 2.40 × 10⁻⁸ cm² · s⁻¹, 20 kDa dextran). In one eye, the result was higher than the 95% CI of the control; in three, it was lower.

conclusions. UES is not caused by reduced scleral hydraulic conductivity, which tends to be higher than expected. Reduced macromolecular diffusion may impede the normal transscleral egress of albumin with subsequent osmotic fluid retention in some, but not all eyes.

Idiopathic uveal effusion syndrome (UES) is an extremely rare disease characterized by suprachoroidal fluid collections and serous retinal detachment (RD), most typically occurring in middle-aged men. UES is sometimes associated with hypermetropia or nanophthalmos, and generally takes a relapsing–remitting course, often with a poor visual outcome. 

The pathogenesis of UES is not certain, but it is possible that the disease occurs secondary to altered scleral structure and permeability. Histologic studies ^{1 2 3} have shown that scleral collagen fibers are typically disorganized, with collections of glycosaminoglycan (GAG)-like material expanding the interfibrillary spaces. Analysis of this material shows it to be mostly proteodermatan sulfate, with some proteochondroitin sulfate. ² A popular theory ² is that these deposits impede transscleral water movement—that is, they reduce hydraulic conductivity (HC), yet proteodermatan sulfate may be expected to bind water rather than exclude it. Another theory is that they result in scleral swelling that compresses the vortex veins, ⁴ compromising choroidal outflow. This theory has led some surgeons to advocate decompression of the vortex veins as a treatment. Although some success has been reported with this technique, ⁴ the most common treatment is full-thickness sclerectomies, as described by Gass and Johnson. ^{5 6} The effectiveness of this surgery is variable, but some patients undoubtedly respond to surgery, and this has been marshaled as evidence that altered scleral permeability is the primary defect in UES. ⁵ This logic may be questioned, as producing a conduit for suprachoroidal outflow would be expected to relieve not only reduced transscleral water and protein movement, but fluid collection due to any other mechanism. 

In the present study, we sought to test the hypothesis ² that UES is caused by reduced scleral HC. Scleral specimens were removed during surgery for UES, and HC measurements were compared to published normal values obtained by using an identical methodology. ⁷ In addition, when there were sufficient scleral samples available, macromolecular diffusion was studied by using fluorescein-labeled dextrans of various molecular weight (MWT), to test the hypothesis of Gass and Johnson ^{5 6} that reduced transscleral protein movement may be responsible for UES. 

Methods

Patient Recruitment and Selection

Because of the extreme rarity of the diagnosis, multicenter recruitment was undertaken via the British and Eire Association of Vitreoretinal Surgeons. Delegates were invited to refer patients or send surgical samples for analysis. Surgery was undertaken at six National Health Service teaching hospitals located in London (Moorfields Eye Hospital and St. Thomas’ Hospital), Manchester, Leeds, Sheffield, and Southampton. Patients were identified before surgery for UES, and the recruiting hospitals were provided with an information pack detailing inclusion and exclusion criteria, a patient consent form, and a pro forma to collect clinical data prospectively. A further form was sent to collect follow-up data on each patient’s postoperative clinical course. The study was approved by the Moorfields Eye Hospital Research Ethics committee and complied with the Declaration of Helsinki, with informed patient consent. 

For inclusion, patients had to have a clinical diagnosis of UES with evidence of both serous RD and ciliochoroidal or choroidal elevation. Patients were excluded if they had a retinal break or other cause of RD, or any other cause of choroidal effusions, such as ocular hypotony or intraocular inflammation. The pro forma requested the following information, if available: age, sex, systemic diagnoses, systemic and ocular medications, refraction, axial length, ultrasound findings, presenting visual acuity (VA), duration of RD and choroidal elevation before surgery, surgical technique, intraoperative examination of the sclera (thickened, rigid, and other observations), and postoperative findings including VA and duration of resolution of the choroidal fluid. The treating vitreoretinal surgeon was also invited to provide a summary of the salient clinical features. 

Surgery

Scleral samples were obtained during sclerectomy surgery to treat UES, using published techniques, ⁶ with sclerectomies sited at or near the equator. The number of sclerectomies (one to four) varied at the discretion of the vitreoretinal surgeon, but in all cases, the sclerectomy depth was full thickness, or deep enough to visualize the choroid and allow fluid to seep out of the eye through the sclera. In some cases, an overlying scleral flap was fashioned to moderate fluid outflow. Primary surgery was defined as an anatomic success if choroidal effusions settled fully after one operation, a partial success if there was a reduction in the choroidal effusion, and unsuccessful if the fluid collections failed to improve within 3 months of surgery, or if more than one operation was required in the same eye. 

HC and Scleral Thickness

Sclerectomy specimens were transferred to the study laboratory by courier in a sterile surgical pot containing saline, placed on ice, in an insulated box. Specimens were not frozen or fixed before testing HC. Surgical specimens typically measured at least 3 by 4 mm, but the amount of available tissue, including the number of specimens, varied. In all cases, priority was given to measurement of HC, and if extra tissue was available, it was used for histology and measurement of transscleral macromolecular diffusion, as described later. 

Scleral specimens were examined under a dissecting microscope to exclude perforating surgical defects. An outer corner of the square scleral specimen was removed, fixed, and used for histologic analysis, and the remaining unfixed specimen was mounted in a Üssing chamber for testing of HC. HC was measured as described previously. ⁷ Briefly, tissue was clamped in the Üssing chamber with the inner scleral surface connected to a water column set at a pressure of 15.7 mm Hg, corresponding to a normal intraocular pressure. In most cases, a 4-mm aperture was used, but some smaller specimens required a modified aperture of 1.5 or 3 mm. Water was allowed to pass across the scleral specimen under this pressure head. As the water passed across the scleral substrate, there was a corresponding descent in the water column. This was measured with a digital micrometer aligned with the lower meniscus, visible within the glass tubing that contained the water column. The volume per unit length of the glass tubing was predetermined with mercury weighed on analytic scales. Hence, the descent of the water column over 24 hours could be used to calculate the HC (the flow of water, per unit surface area, and per unit pressure). 

Tissue for light microscopy was fixed for at least 1 hour in 2.5% glutaraldehyde in 0.1 M sodium cacodylate-HCl buffer containing 10 g/L calcium chloride. Tissue was rinsed in sucrose and then postfixed in 2% osmium tetroxide in 0.2 M sodium cacodylate buffer. The samples were rinsed in 20% ethanol, and then dehydrated in a graded series of ethanol. Sclera was transferred to a 50:50 solution of ethanol and propylene oxide (1,2-epoxypropane), 100% propylene oxide, and epoxy resin (30 g Araldite CY212, 25 g dodecenylsuccinic anhydride hardener 964, 0.9 mL 2,4,6-tri (dimethylaminoethyl) phenol; Agar, Cambridge, UK) diluted 50:50 with propylene oxide that was then evaporated overnight. Tissue was transferred to fresh 100% epoxy resin for 8 hours, poured into to rubber molds, and polymerized for 24 hours at 60°C. Tissue blocks were cut with a razor blade, and then polished on a microtome (Huxley Instruments, Huxley, UK). 

Semithin (1 μm) sections were cut with a glass knife and transferred to plain microscope slides containing a film of distilled water. If necessary, the sections were flattened by using trichloroethylene vapor. These were dried on a hot plate for 20 minutes and then stained with filtered toluidine blue for 30 to 60 seconds. The slides were then rinsed with 60% ethanol and allowed to dry on a hot plate. Scleral thickness was measured with a sliding graticule on a light microscope, taking the mean of 15 randomly selected histologic sections. In two patients (patient 4 and 5) scleral thickness was measured with a micrometer (unfixed tissue), because of insufficient availability of tissue for histology. Although our purpose was not to study scleral ultrastructure, in two patients (9 and 10) transmission electron microscopy (TEM) and additional staining were undertaken by the clinical laboratories where the patient was treated and were included in the results. Histologic sections were viewed in a masked fashion, before any attempts to correlate with experimental results, but no formal grading system was applied. 

Care was taken to prevent tissue dehydration during HC testing and to maintain the orientation of the scleral specimens during histologic processing. Our previous experiments established the integrity of scleral tissue clamped in the Üssing chamber, ⁷ and other investigators have shown that fixed specimens give a good estimate of in vivo (unfixed) scleral thickness. ⁸  

HC was corrected for the scleral thickness of each specimen to give specific HC, or K. This value (K) therefore reflects the intrinsic HC of scleral tissue, independent of path length (thickness). Using K, it was possible to compare the HC of any partial-thickness surgical specimens, with our normal values obtained in full-thickness sclera from cadaveric eyes. ⁷ The scleral specimens in the previous cadaver study were taken from the same topographic region—namely, equatorial or pre-equatorial sclera. K was calculated as reported previously ⁹ and as shown in Appendix A. 

Dextran Diffusion

In patients in whom there were sufficient intact samples to perform additional experiments, an assessment of transscleral macromolecular diffusion was made. The rate of diffusion across the scleral substrate was determined for fluorescein isothiocyanate (FITC)-labeled dextrans of different molecular weight (MWT). Experiments were based on previously described techniques used to determine the retinal MWT exclusion limit. ¹⁰ Specimens were not frozen or fixed before diffusion studies. 

Tissue was carefully inspected under a dissecting microscope to ensure that there were no perforating defects. The scleral specimen was then clamped in a modified Üssing chamber with the inner scleral surface facing a hemichamber containing 0.412 mM FITC-dextran (Sigma-Aldrich, Poole, UK) dissolved in phosphate-buffered saline (PBS) with added penicillin (100,000 U/L), streptomycin (100 mg/L), and amphotericin B (250 μg/L). The second hemichamber contained PBS with the same antibiotics and antimycotic. In most cases the interchamber aperture measured 4 mm, but some smaller specimens required a modified aperture of 1.5 or 3 mm. The fluid volume in each hemichamber was kept identical throughout the mounting procedure and during subsequent fluid removal to prevent a hydrostatic pressure gradient from developing across the sample. The locking screws that held the assembly together were tightened with a torque-range screwdriver (RS Components, Northants, UK). The minimum torque needed to prevent leakage around the interchamber aperture was applied (≈30 cN/m). The tops of the chambers were sealed with insulation tape to prevent evaporation and a small, glass-encased magnetic stirrer was inserted into the bottom of each hemichamber. Experiments were conducted at 20°C, protected from ambient illumination. The integrity of tissue clamped in the Üssing chamber has been established previously. ^{7 10}  

Experiments were initiated with the lowest MWT FITC-dextran (7–20 kDa). Once FITC was visible in the second hemichamber, samples were removed, and the absorbance at 490 nm was determined by using a spectrophotometer (UV-160; Shimadzu, Kyoto, Japan). The change in FITC-dextran concentration over time was then calculated from predetermined standard curves for each dextran. After each experiment the fluid in each hemichamber was replaced with PBS, and any remaining dextran was allowed to diffuse out of the tissue and into the hemichambers. The fluids were serially exchanged until no FITC-dextran was detected. The experiment was then repeated with the next lowest MWT dextran, and so on. When more than one scleral specimen was available, the mean value was reported. 

Results were given as a diffusion coefficient (D). This coefficient incorporates path length (scleral thickness) and therefore allows comparison with results in full-thickness, disease-free human sclera. The calculation of D is given in Appendix B, and is as defined in recent studies of transscleral diffusion. ¹¹  

To allow comparison with other reports in the literature, we also estimated D at 37°C, as calculated by other groups such as Boubriak et al., ¹² and as shown in Appendix C. Normal data were obtained from disease-free cadaveric eyes from the UK Tissue Transplant Service (Bristol, UK), with samples taken from the same topographic region, near to or just anterior to the equator. 

Results

Clinical Data

Follow-up forms were completed for all 10 patients, 4 of whom had surgery in both eyes. Mean follow-up from the most recent operation was 22 months (range, 2–60) and mean age was 63 years (range, 43–79). Twelve of 14 eyes were hypermetropic with a mean spherical equivalent +4.7 D. Preoperative VA ranged from 6/6 to hand motions, with the median VA improving from 0.20 (6/30) to 0.33 (6/18) after surgery. Nine eyes showed visual improvement, three worsened, and two remained unchanged. Final VA ranged from 6/6 to hand motions. As defined in the Methods section, initial surgery was anatomically successful in 7 of 14 eyes, partially successful in 2, and failed in 5. Four eyes required more than one operation, with a total of 19 operations in 14 eyes. Clinical data are summarized in Table 1 , with examples shown in Figures 1 and 2 . A clinical synopsis of each patient is presented in Supplementary Table S1. 

HC and Scleral Thickness

HC was measured in all patients. The mean (±SD) HC was 43.4 ± 58.0 × 10⁻¹⁰ cm² · s⁻¹ · Pa⁻¹, and the mean thickness of the excised scleral specimens was 585 ± 309 μm. Mean K at 20°C was 23.9 ± 27.6 × 10⁻¹⁴ cm². Individual results are shown in Table 2 . Results were compared to published values obtained in disease-free cadaveric eyes, by using the same methodology. ⁷ This comparison showed that patients with UES had a higher mean K (23.9 ± 27.6 vs. 5.8 ± 3.9 × 10⁻¹⁴ cm²; Fig. 3 ). Although this difference was not quite significant (Mann-Whitney P = 0.0684), patients 8, 9, and 10 were 8 to 18 SD above the control mean, and well outside the 95% confidence interval (Fig. 4) . 

Samples obtained from patients with UES were transported on the day of surgery or the next morning, and consequently experiments were undertaken more quickly than in eye bank eyes that were used as control specimens. Experiments in the UES group were concluded within 2 to 5 days of surgery. Previous analysis of eye bank eyes failed to demonstrate a significant alteration in HC with time, at least not over the timeframe used in those experiments. ⁷  

Dextran Diffusion

There were sufficient samples to determine D in five patients, with individual results shown in Table 3and combined results shown in Figure 5 . As can be seen in the table, there was a complete data set for the 20-kDa dextran, and therefore control data were obtained by using this dextran. Aged-matched, disease-free cadaveric eyes showed a mean D (±SD) of 6.14 ± 2.40 × 10⁻⁸ cm² · s⁻¹, similar to the mean value in eyes with UES (5.69 ± 5.35 × 10⁻⁸ cm² · s⁻¹; Fig. 6 ). Despite this overall similarity, individual results varied considerably, with three patients below the 95% CI of normal data and one above, as shown in Figure 7 . The control data in Figure 7show a significant, age-related decline in D with age. 

Histology

In 8 of 10 patients (all except patients 4 and 5) there were histologic specimens of sufficient quantity and quality to assess scleral microstructure. Light microscopy showed features consistent with previous reports, ^{1 3} with disorganization of the collagen bundles, and collections of amorphous, acellular deposits interspersed between fibers. There was no clear pattern between the extent of these deposits and the K, D, or clinical presentation, other than the fact that the two most hypermetropic patients (patients 3 and 7) had more deposits than others in the series. There was very mild hypercellularity in some of the samples (patients 2, 3, and possibly 8), but no manifest inflammatory change. Representative histology is shown in Figure 8 . TEM results in patients 9 and 10 (undertaken by the treating hospital) are given in Supplementary Table S1. 

Discussion

In this study, we investigated the transscleral flow of water in patients with UES and found that the intrinsic HC of scleral tissue was similar to previously published results in control eyes of a similar age. ⁷ This finding has several implications. 

First, the hypothesis that reduced scleral water permeability is responsible for idiopathic UES can be rejected, at least in this patient group. This conclusion is consistent with results in our previous experiments in normal sclera. ⁷ These showed that the potential flow of water across the sclera is considerably more than occurs in vivo, with the capacity for transscleral water movement approximately two to three times estimates of uveoscleral outflow. The fact that scleral outflow is not fully realized in vivo suggests that there are barriers to aqueous movement into the suprachoroidal space, as would be expected given that aqueous has to pass through stromal tissue at the drainage angle. There is evidence that the ciliary muscle acts as a potential barrier, ¹³ as does the presence of an anatomic “compact zone” at the ora serrata. ¹⁴ That suprachoroidal pressure is below intraocular pressure ¹⁵ further suggests that aqueous movement is at least partially impeded. When these barriers are overcome, as with a cyclodialysis cleft, the capacity for transscleral water movement is realized and intraocular pressure can drop precipitously. Taken together, these facts confirm that the sclera has considerable spare capacity to remove suprachoroidal water. Therefore, there would have to be a large reduction in HC to produce the suprachoroidal fluid collections seen in UES, if this were the only causative mechanism. 

Second, the increased scleral thickness that can occur in UES is unlikely to alter HC enough to explain the effusions associated with this condition. From the equation shown in Appendix A it follows that flow across the sclera (Q) reduces with increasing scleral thickness (L), but even if scleral thickness in patients with UES was increased fourfold, flow would still be similar to that of the controls, as the mean K obtained in this study was approximately four times higher than control data obtained using identical methodology. The posterior pole ultrasound measurements taken on patients 1, 2, 4, 9, and 10 showed scleral thickness measurements of 1.25 to 2.9 mm. Comparable normal values are 0.8 to 1.0 mm, ⁸ suggesting UES scleral thickening is not enough to reduce HC below normal values. Further, the increased interfibrillary spaces seen histologically in UES may increase water permeability, in the same way that scleral hydration is known to expand the interfibrillary spaces and increase permeability. ¹⁶ It is interesting therefore that patient 9 had both the highest HC and the thickest scleral sample. Taken together, these findings suggest that increased scleral thickness is unlikely to significantly impede scleral HC relative to normal subjects, if anything there was a trend for increased HC. 

This might not be the case, however, if the scleral composition changed from the inner to outer sclera, such that the innermost sclera had a significantly lower HC—that is, if our scleral specimens were not representative of full-thickness sclera. However, our histologic studies and those of others ^{1 3} do not suggest this to be the case, and indeed most patients had full-thickness or near to full-thickness scleral specimens removed at the time of surgery. 

An alternative hypotheses is that fluid is retained in the choroidal space by osmotic forces. Albumin (66 kDa) is the most important serum protein in relation to osmotic pressure. Bill ¹⁷ demonstrated that labeled albumin injected into the rabbit suprachoroidal space leaves primarily via a transscleral route, and because proteins can traverse the sclera, the colloid osmotic pressure in the suprachoroidal space is effectively zero. ¹⁸ If transscleral macromolecular diffusion were reduced, then this would tend to retain protein and thus fluid in the suprachoroidal space, particularly if the fenestrated capillaries of the choroid did not have the capacity to remove retained protein across the sclera or if their normal capacity was reduced because their exit across swollen sclera was compromised. 

Reduced diffusion might be expected as studies show increased deposition of glycosaminoglycan (GAG)-like material in the scleral substrate of patient with UES, ³ and it is known that increased concentrations of GAGs can reduce diffusion across interstitial tissue, ⁹ including human sclera. ¹² Further, experimental removal of GAGs increases diffusion. ¹² It may therefore be predicted that the amorphous material in our patients and in those in other studies ^{1 3} has a similar effect. 

Three of five patients in this series had reduced diffusion of a 20-kDa dextran, when compared with that in control subjects. The trendline in Figure 5shows the expected ^{12 19 20} linear relationship between the rate of diffusion and log(MWT), so that it appears diffusion may be affected equally across a range of high MWT molecules. This result supports the hypothesis that altered transscleral macromolecular diffusion may cause or contribute to UES. However, mean values were similar to those in the control subjects, and two of five cases had normal or high rates of diffusion. This indicates that impaired permeability to high MWT molecules cannot explain the pathogenesis of UES in all cases. 

The diffusion studies presented in this article were undertaken at near to atmospheric pressure, distinct from the HC studies, in which water movement was examined under a hydrostatic pressure head (set at intraocular pressure). It is possible therefore that in vivo diffusion is reduced if the intraocular pressure compresses the inner scleral fibers; however, this compression would affect both UES specimens and controls equally. In addition, other researchers’ work ^{20 21} suggests that this effect, if present, is likely to be small in a disease with typically normal intraocular pressure. 

Until now, hypotheses on the pathogenesis of UES were largely untested, and experimental studies were limited to histologic observations. In this respect, this is the first experimental study of UES. Another strength of this prospective study is that we used a validated methodology to make a direct and quantitative assessment of scleral permeability, with data that may contribute usefully to our understanding of UES. Given the rarity of UES, and in comparison to most series, this is a large study. However, with only 10 patients there is inevitably a spread of results and findings may not apply to all patients with this condition. Further, surgical tissue acquisition is less standardized than in cadaveric eyes, and variability in the size of the scleral specimens can make tissue mounting difficult, with the risk of artifact and variable results. 

In summary, we found that transscleral hydrostatic water movement in patients with UES was similar to that in healthy control subjects, despite manifest alterations in scleral structure. The commonly reported hypothesis that altered scleral HC is responsible for UES can be rejected in this patient group. Macromolecular diffusion may be reduced in some patients, and this would tend to retain albumin in the suprachoroidal space with a resulting osmotic retention of fluid, assuming that protein could not be cleared by the fenestrated choroidal vessels. However, not all patients showed reduced macromolecular diffusion. In these cases, secondary constriction of transscleral emissaria remains a viable explanation for the choroidal fluid collections seen in UES. 

Appendix 1

Specific hydraulic conductivity (K), in square centimeters, is defined as  

\[K{=}\ \frac{Q\ {\cdot}\ {\mu}\ {\cdot}\ L}{A\ {\cdot}\ {\Delta}P}\ ,\]

where Q is flow, determined from the descent of the water column in centimeters, multiplied by the constant of the individual tube (volume per unit length in square centimeters), divided by time in seconds (cubic centimeters per second); μ is viscosity of water in experimental conditions (1.003 × 10⁻³ Pa · s at 20°C), or at 37°C (0.692 × 10⁻³ Pa · s); L is path length (scleral thickness in centimeters); A is the surface area of aperture (in square centimeters); and ΔP is the mean pressure head over the duration of the experiment (in pascals). 

Appendix 2

The calculation of the diffusion coefficient (D) (in square centimeters per second) is based on Fick’s first law:  

\[D{=}\ \frac{R\ {\cdot}\ L}{A\ {\cdot}\ C}\ ,\]

where R is diffusion flux (in nanomoles per second); L is the scleral thickness (in centimeters); A is the surface area of interchamber aperture (in square centimeters); C is the concentration gradient. As the concentration in the receiver chamber was typically approximately 1% of that in the donor chamber, the concentration in the donor chamber was taken as the concentration gradient, in nanomoles per cubic centimeters. 

Appendix 3

Experiments were conducted at 20°C. Einstein’s relationship was used to adjust the diffusion coefficient, to allow comparison with other researchers’ results:  

\[D_{37}{=}D_{20}(T_{37}/T_{20})({\mu}_{20}/{\mu}_{37}),\]

where D ₂₀ and D ₃₇ are diffusion coefficients at 20°C and 37°C (in square centimeter per second); T ₂₅ and T ₃₇ are absolute temperatures in K at 25°C and 37°C; μ₂₀ and μ₃₇ are viscosity of water at 20°C and 37°C (see Appendix A for values). 

 Submitted for publication March 5, 2008; revised May 4 and 26, 2008; accepted August 19, 2008.

 Disclosure: T.L. Jackson, None; A. Hussain, None; A.M.S. Morley, None; P.M. Sullivan, None; A. Hodgetts, None; A. El-Osta, None; J. Hillenkamp, None; S.J. Charles, None; R. Sheard, None; T.H. Williamson, None; A. Kumar, None; D.A.H. Laidlaw, None; W.H. Woon, None; M.J. Costen, None; A.J. Luff, None; J. Marshall, None

 The publication costs of this article were defrayed in part by page charge payment. This article must therefore be marked “advertisement” in accordance with 18 U.S.C. §1734 solely to indicate this fact.

 

Table 1.

 

Clinical Data

Patient Number, Sex, and Age in Years	Axial Length (mm)	Spherical Equivalent (D)	Thickened Sclera Noted during Surgery (Ultrasound Measurement if Available, in mm)	Preoperative VA	Number Of Sclerectomies (Number of Operations)	Anatomic Result, †	Final VA (Follow-up in Months)
1. M79	22.0	NR	No (1.25)	HMs	3 (1)	Success	6/36 (7)
2. M67	23.0 OS	+1.25 OD	Yes OU (2.1 mm OS)	6/36 OS	OS 4(1)	Partial OS	6/12 OD (5)
	NR OD	+0.90 OS		CF OD	OD 7(2)	Failed OD	6/18 OS (13)
3. M43	19.2 OD	+7.50 OD	Yes (OU)	6/12 OD	2 (1) OD	Success OD	6/18 OD (60)
	19.5 OS	+10.00 OS		6/24 OS	2 (2) OS	Failed OS	6/18 OS (4)
4. M73	21.6	+3.25	No (1.4 mm)	6/24	4 (1)	Partial	6/60 (32)
5. F57	22.5	0.0	Yes	6/24	4 (2)	Failed	HMs (6)
6. F66	20.5	+4.50	Yes	HMs	2 (1)	Success	HMs (45)
7. M57	18.47	+13.00	Yes	1/24	6 (2)	Failed	6/12 (12)
8. F72	NR	+2.00 OD	Yes OU	HMs OD	1 (1) OD	Success OD	6/24 OD (54)
		+2.00 OS		6/12 OD	2 (1) OS	Success OS	6/9 OS (24)
9. F75	21.8	+3.75	Yes (1.5 mm OU)	HMs	2 (2)	Failed	2/60 (3)
10. F44	19.76 OD	+6.00 OD	Yes (2.5 mm) OD	6/9 OD	2 (1)	Success (OU)	6/5 (2)
	19.16 OS	+7.25 OS	Yes (2.9 mm) OS	6/6 OS	2 (1)		6/6 (36)

 HM, hand motions vision; NR, not recorded.

*  Additional clinical information is given in Supplementary Table S1.

†  The definition of anatomic success or failure is as given in the Surgery subsection of Methods.

Figure 1.

 

B-mode ultrasound scan of patient 1, showing choroidal elevation and a bullous, inferior retinal detachment.

Figure 2.

 

Top: peripheral fundus photograph of patient 2 before left eye surgery, with a large choroidal effusion; bottom: photograph taken after surgery and after resolution of the effusion. The fundus has a mottled appearance with areas of both hyper- and hypopigmentation throughout.

Table 2.

 

Hydraulic Conductivity and Scleral Thickness

Patient	Hydraulic Conductivity at 20°C (×10−10 cm · s−1Pa−1)	Scleral Thickness (μm)	K at 20°C (×10−14 cm2)	K at 37°C (×10−14 cm2)
1	7.7	523	4.04	2.78
2	15.6	543	8.50	5.84
3	16.6	1041	17.33	11.92
4	10.0	380	3.81	2.63
5	16.6	400	6.60	4.59
6	46.5	350	16.32	11.23
7	2.0	425	0.85	0.59
8	60.6	610	37.08	25.51
9	62	1240	77.11	53.2
10	196	340	67.00	46.1
Mean	43.4	585	23.86	16.44

Figure 3.

 

Mean (±SD) scleral specific HC (K) in 16 healthy control adult eyes (mean age, 63 ± 8 years), obtained from previously published normal data. ⁷ The result is compared to the mean K in 10 scleral specimens from patients with UES (mean age, 63 ± 13 years). The difference was not quite significant (P = 0.0684).

Figure 4.

 

Previously published scleral-specific HC (K) data obtained from 18 disease-free cadaveric eyes over a range of ages (○). ⁷ Solid gray line: significant reduction in K with age (linear regression, y = −0.20x + 19.74; R ² = 0.24; P = 0.039); dotted lines: the 95% confidence interval. These data are compared with the present data set (▪), obtained from surgical samples of patients with UES. Most patients had HC similar to that of the control subjects; however, three patients (8, 9, and 10) had higher values than normal.

Table 3.

 

Macromolecular Diffusion

Dextran Size (kDa)	Diffusion Coefficient at 20°C (cm2 · s−1)	Diffusion Coefficient at 37°C (cm2 · s−1)
Patient 1
0.376 (sodium fluorescein)	2.45 × 10−7	3.76 × 10−7
4.4	1.35 × 10−7	2.07 × 10−7
9.5	1.65 × 10−8	2.53 × 10−8
19.5	2.64 × 10−8	4.05 × 10−8
42	4.39 × 10−8	6.73 × 10−8
77	6.97 × 10−9	10.69 × 10−8
Patient 4
12	3.42 × 10−8	5.25 × 10−8
21.2	1.51 × 10−8	2.32 × 10−8
38.5	6.86 × 10−9	10.52 × 10−9
70.5	3.11 × 10−9	4.77 × 10−9
Patient 5
21.2	8.33 × 10−8	12.77 × 10−8
38.2	3.95 × 10−8	6.06 × 10−8
70.5	2.54 × 10−8	3.90 × 10−8
Patient 9
21.2	1.39 × 10−7	2.13 × 10−7
42	4.83 × 10−8	7.41 × 10−8
77	2.61 × 10−8	4.00 × 10−8
Patient 10
20	2.06 × 10−8	3.16 × 10−8

Figure 5.

 

Rate of dextran diffusion across scleral specimens related to MWT, with a log scale. Data were pooled from the five patients from whom sufficient scleral specimens were obtained to undertake dextran diffusion studies (patients 1, 4, 5, 9, and 10, as indicated). Linear regression showed an inverse relationship (P = 0.035) between diffusion coefficient and log(MWT).

Figure 6.

 

Mean (±SD) diffusion coefficient of the 20-kDa dextran. The seven healthy control eyes came from donors with a mean age (±SD) of 68 ± 19 years, and the result in this group is compared to that in five patients with UES (mean age, 66 ± 15 years).

Figure 7.

 

(○) The diffusion coefficient of healthy control sclera in seven eyes. Linear regression in this group showed a statistically significant (y = −7.5518x + 114.53; R ² = 0.8779; P = 0.0019) reduction in transscleral diffusion with increasing age. (▪) Samples from the five patients with UES. Three of these samples had a diffusion coefficient below the 95% confidence interval of the control group (dotted gray lines), and one had a diffusion coefficient above this level.

Figure 8.

 

Scleral sections from patient 2. (A) Positive staining of the scleral fibers with Alcian blue. Elastin van Geison stain (B) showed relatively little elastin, possibly some fragmentation of the collagen bundles, but appearances were generally nonspecific. The apparent spacing between the fibrils is artifactual. (C) Photomicrograph of tissue obtained from patient 3 and stained with Toluidine blue. Amorphous deposits were seen interspersed between the collagen bundles.

Supplementary Materials

Supplementary Table S1 - (PDF)