This year's glaucoma research covers a broad range of topics, from medications and surgery to associated factors, genetic influences and new approaches to diagnosis. (Unless otherwise noted, studies were conducted without commercial support.)

 

Factors That Impact Glaucoma

Researchers at the University of Michigan conducted a study to determine whether an association exists between the various components of metabolic syndromes (diabetes mellitus, arterial hypertension, hyperlipidemia and obesity) and open-angle glaucoma in a large, diverse group of individuals throughout the United States.

Researchers identified all beneficiaries age ≥40 years continuously enrolled in a large managed-care network who made at least one visit to an eye-care provider between 2001 and 2007. ICD-9 billing codes identified individuals with OAG and those with components of metabolic syndrome. Cox regression was used to determine the risk of developing OAG in patients with components or combinations of components of metabolic syndromes. Adjustments were made for sociodemographic factors, systemic medical conditions and other ocular diseases.

Of the 2,182,315 enrollees who met inclusion criteria, 54,558 (2.5 percent) had OAG. After adjustment for confounding factors:

   •Those with HTN (HR=1.35) or DM (HR=1.17) had an increased hazard of developing OAG relative to persons with neither of these conditions. Those who had HTN and DM in combination also had an increased risk (HR=1.48).

   • The presence of hyperlipidemia in combination with HTN (HR=1.13) or with DM (HR=1.09) produced hazard ratios of lesser magnitude.

   • Persons with hyperlipidemia alone had a 5-percent decreased hazard of OAG (HR=0.95) compared with those who had no components of metabolic syndrome.

The authors note that determining whether hyperlipidemia or the medications used to treat this condition reduce the hazard of OAG could lead to improved OAG treatment strategies.3974

Researchers at the Tokyo Metro-politan Police Hospital in Japan conducted a study to determine how age, gender, refractive error, mean 24-hour IOP, systolic blood pressure, diastolic blood pressure and pulse rate are associated with 24-hour IOP fluctuation. They examined 128 eyes of 128 patients with previously unoperated NTG (74 males and 54 females, mean age 53.7 ±13.5 years). Patients who had previously been receiving topical ocular hypotensive agents were asked to stop their use for ≥four weeks. The patients were then hospitalized for 24 hours to measure IOP.

The data showed that NTG patients with a higher degree of myopia had bigger 24-hour fluctuation of IOP. Fluctuation was significantly negatively correlated with age (β=-0.008, r2=-0.035, p=0.034) and refractive error (β=-0.033 ,r2=0.046, p=0.015).563

Noting that Helicobacter pylori infection and NTG are both common among Koreans, researchers at three centers in Korea conducted a study to see whether the two were statistically associated. They enrolled 100 consecutive NTG patients from an outpatient clinic (Group I) and reviewed medical records of 104 patients newly diagnosed with NTG (Group II) as well as 1,116 controls. Serum samples from all subjects were analyzed for the presence of H. pylori antibodies.

The NTG patients had significantly more positive serologic results than did the normal controls (Group I: 72 percent vs. 28 percent, p=0.018; Group II: 73.1 percent vs. 59.8 percent, p=0.008, OR=1.83). No significant association was found between clinical characteristics and a positive serologic result in NTG patients. The authors note that H. pylori may play a role in the development or progression of NTG as a secondary aggravating factor, or could be a primary cause.190

 

Glaucoma Medications

Researchers at two universities in Ontario, Canada, compared the bottle design, viscosity, surface tension and drop volume of brand-name and generic topical glaucoma drops in Canada and the United States, to determine whether the dosage received by users was the same. The data revealed statistically significant differences in viscosity between brand name and generics; in surface tension; and in the outer orifice diameter of the dispensing bottles. Overall, the generics Timolol Maleate EX (Canada) and Timolol GFS (U.S.) deliver approximately three-fifths and two-thirds, respectively, of the daily therapeutic dosage delivered by the same country's Timoptic XE. (See chart.)194

Researchers at the Konkuk Uni-versity Hospital in Seoul, Korea, conducted a retrospective study to evaluate the long-term effect of latanoprost on central corneal thickness in patients with normal tension glaucoma. They included 128 patients with newly diagnosed NTG and no previous topical treatment, and 38 patients with suspected glaucoma (suspicious disc, normal visual field and IOP less than 21 mmHg) as the control group. There were no significant differences in mean sex, age, baseline IOP or CCT between the groups. The experimental group received latanoprost 0.005% monotherapy; patients were followed for more than 24 months. CCT measurements before treatment and 24 months after treatment were analyzed with ultrasound pachymetry.

A statistically significant reduction in mean CCT was observed in the latanoprost group (535.5 ±37.9 vs. 530.1 ±36.4 µm, p<0.001) but not in the control group (543.1 ±40.2 vs. 542.6 ±37.0µm, p=0.786). The authors note that CCT variations resulting from medical treatment could affect IOP measurement in NTG patients.179

 

Imaging

Noting the increased use of spectral domain optical coherence tomography in glaucoma assessment, researchers at multiple centers in the Netherlands investigated the potential problem of differences in retinal reflectivity between healthy and glaucomatous eyes, which could affect the ability of SD-OCT to accurately determine the thickness of the retinal nerve fiber layer. They scanned the peripapillary areas of eight healthy and eight glaucomatous eyes in four quadrants at fixed locations with the Spectralis OCT (Heidelberg) and determined the reported reflectivity of each manually segmented retinal layer.

The mean reflectivity of the RNFL was significantly lower in glaucomatous eyes than in healthy eyes: 0.506 (SE 0.015) vs. 0.619 (SE 0.015) (p< 0.01). Also, the contrast between the RNFL and the GCC was significantly lower in glaucomatous eyes than in healthy eyes: 0.123 (SE 0.011) vs. 0.232 (SE 0.011) (p<0.01). They conclude that the differences in contrast and reflectivity may impair accuracy and limit the use of SD-OCT in glaucoma management.212

Researchers at the University of Iowa and Iowa Veterans Admin-istration Medical Center conducted a study to validate an algorithm they've created that's able to generate depth maps from stereo color images of the optic disc—a widely used, low-cost tool that has lacked a means of objective assessment. For validation they compared 30 depth maps and SD-OCT scans of the optic disc taken on the same day from both eyes of glaucoma patients. (See image, above.)

The root mean squared difference between normalized structures was 0. 1592. The automatically computed depth maps based on stereo color images of the optic disc faithfully reproduced the shape of the ONH generated by SD-OCT. The authors propose this method as a low-cost alternative to SD-OCT for diagnosis and management of glaucoma and other optic neuropathies.1774

 

Glaucoma and Daily Living

Researchers at multiple centers in the United Kingdom examined the relationship between combined physical activity (at work and leisure) and IOP, measured in 2,613 men and women aged 48 to 89 participating in the EPIC-Norfolk study between 2006 and 2009. Individuals were categorized as inactive, moderately inactive, moderately active or active. Measurements taken included: IOP (using the Reichert Ocular Response Analyzer and Goldmann tonometer); blood pressure; body mass index; and hip and waist circumferences.

The data showed that the highest level of physical activity was associated with lower IOP, lower BMI, lower systolic blood pressure and smaller waist/hip ratio. Lower IOP was also associated with lower BMI and lower SBP. Furthermore, the most active participants' IOP was an average of 0.5 mmHg lower than the IOP of the most inactive participants. (However, adjusting for BMI, SBP and WHR attenuated the difference in mean IOP between active and inactive participants.)

The authors note that physical activity and weight control may offer a safe, simple method of reducing IOP.3977

Researchers in the National University Health System in Singapore conducted a prospective observational case series to determine whether head postures adopted during sleep have an effect on IOP.

Twenty-three adult patients without glaucoma, ocular hypertension or any other conditions associated with raised IOP, who were able to lie recumbent in the supine and left lateral decubitus position, were recruited from general ophthalmology out-patient clinics. Mean age was 54 ±16 years. Ocular diagnoses included cataracts (30 percent), dry eyes (26 percent), blepharitis (22 percent), vitreous syneresis (13 percent) and other (9 percent). The IOP in the left eye of each patient was measured in the supine position using a Tono-Pen XL, immediately followed by a second measurement obtained in the left lateral decubitus head position. Measurements in both positions were obtained with the patient lying first on a soft pillow, then on a hard pillow.

The mean IOP in supine head position was 14.1 ±3.3 mmHg; the mean IOP in left lateral head position was 17.1 ±3.4 mmHg—a statistically significant change (p<0.001). The shift from soft to hard pillow made no difference in IOP in either head position. The authors note that this possible increase in IOP during sleep could lead to progression of glaucomatous optic neuropathy despite target IOP being achieved on examination at routine follow-up.558

Researchers at the University of Alabama at Birmingham conducted a study to evaluate the association between specific visual field characteristics and motor vehicle collision (MVC) involvement in 438 glaucoma patients (mean age: 72.8 years; 55.5 percent female, 60.5 percent white, the remainder black). All subjects had been seen at least once between January, 1994 and December 1995 in one of three university-affiliated ophthalmology and optometry practices. Investigators obtained information on use of glaucoma medications, BCVA in each eye, and visual fields (either a 30-2 or 24-2 test) in each eye. Binocular visual field measures (created by averaging individual eye measures) were stratified into nine regions, and average threshold (TH), total deviation (TD) and pattern de-viation (PD) values were obtained. Finally, demographic, driving, health and other information was obtained via a telephone survey, and information regarding all MVCs that occurred between January 1994 and June 2000 was obtained from the Alabama Department of Public Safety.

Over the entire visual field TH, TD and PD impairment was associated with a non-significant 50-percent, 44-percent and 68-percent increase in the rate of at-fault MVCs, respectively. However, once the visual field was separated into regions, significantly increased rates were observed for certain locations (i.e., the inferior region) and were limited to specific measures (i.e., TD and PD). The authors note that this data might provide clinicians with a set of markers to identify glaucoma patients at risk of motor vehicle collision involvement.3986

Improving Glaucoma Surgery

One of this year's abstracts presents two-year follow-up results from the Ahmed vs. Baerveldt Study, a multicenter, randomized clinical trial comparing the Ahmed-FP7 valve to the Baerveldt-350 implant for treating refractory glaucoma. In the study, 250 patients at least 18 years old with uncontrolled glaucoma refractory to medical, laser and surgical therapy were randomized to receive either an Ahmed-FP7 valve (n=129) or a Baerveldt-350 implant (n=121) using standardized surgical technique.

Results at two years included:

   • Mean IOP was 14.4 ±6.7 mmHg in the Ahmed group and 13.7±4.9 mmHg in the Baerveldt group (p=0.58).

   • Mean number of glaucoma medications was 1.43 ±1.49 in the Ahmed group and 0.77 ±1.06 in the Baerveldt group (p=0.012).

   • The cumulative probability of failure was 48.4 percent in the Ahmed group and 44.2 percent in the Baerveldt group (p=0.78).

   • 22 patients (17.1 percent) in the Ahmed group had complications which required intervention, compared to 31 patients (25.6 percent) in the Baerveldt group (p=0.098).

Overall, at two years the two groups had similar success rates, IOPs and VAs, and required a similar number of interventions to treat complications of surgery. However, the Ahmed group required a greater mean number of glaucoma medications than the Baerveldt group.4443

Researchers at two universities in Ontario, Canada, conducted a retrospective chart review of 114 eyes of 99 glaucoma patients who had undergone bleb needling revision with mi-tomycin-C post-trabeculectomy, to determine the factors predicting the failure of this procedure. After sub-conjunctival injection of 0.02 ml mitomycin-C (0.2 mg/ml), a 30-ga. needle was used to puncture the cyst (for conjunctival or Tenon's cystic blebs), or undermine the scleral flap (for fibrotic blebs). Patients were followed for at least 12 months after needling. The success and failure groups were compared at one, six and 12 months.

The data showed:

   • IOP immediately pre- and post-needling was 24.1 ±5.8 mmHg and 10.4 ±5.6 mmHg, respectively (p<0.001).

   • Success of needling was 63 percent at one month, 53 percent at six months and 41 percent at 12 months.

   • Factors correlating to needling failure included: non-Caucasian race (p=0.003, RR=1.8); re-op trabeculectomy (p=0.05, RR=1.6); history of previous needling (p<0.001, RR=2.1); IOP >25 mmHg before needling (p<0.001, RR=2.0); IOP >12 mmHg after needling (p<0.001, RR=2.0); and resumption of glaucoma meds post-needling (p<0.001, RR=2.2).611

Researchers at the University of Virginia conducted a retrospective, interventional, noncomparative case series to describe surgical applications and clinical outcomes of 10 consecutive patients (mean age 47) who underwent primary or secondary glaucoma surgery in one eye using sterile, irradiated cornea allograft.

The grafts were used to cover a primary glaucoma tube shunt in three patients; to cover a revised or repositioned glaucoma tube shunt in five patients; for tectonic support of a revised trabeculectomy flap in one patient; and to cover subconjunctival prolene sutures following a surgical iridoplasty in one patient. Allografts were either full-thickness (n=5) or split-thickness (n=5). Mean follow-up was 10 months (r: 8.2 to 12.6 months).

No case required revision of the allograft. The allografts demonstrated good biocompatibility in all 10 cases during the follow-up period, as evidenced by maintained clarity of the corneal tissue, absence of clinical signs of immunologic rejection or ocular inflammation, and integrity of the sterile cornea and the conjunctiva over the graft. Cosmetic results were excellent, and there were no complications during the follow-up period.600

 

The Family Connection

Researchers at Fudan University in Shanghai, China, conducted a study to investigate the influence of different elements of family history on the presence and severity of primary angle-closure glaucoma and primary open-angle glaucoma among Chinese patients. They enrolled 332 PAC pa-tients, 228 POAG patients and 193 cataract patients without glaucoma as the controls. All participants underwent a comprehensive eye examination and were asked to finish a questionnaire.

The data showed:

   • Of the 332 PAC patients, 83 (25 percent) had a family history of glaucoma. Characteristic-adjusted odds ratio of family history for PAC was 4.82; OR for severity was 1.61.

   • The presence of first-degree relatives with glaucoma contributed to the odds of having PAC (OR 5.09), as did the number of first-degree relatives (OR 4.71). However, among first-de-gree relatives, only parents contributed to the odds of having PAC (OR 8.76)—not siblings or offspring.

   • Of the 228 POAG patients, 49 (21.49 percent) had a family history of glaucoma (OR 8.38); OR for severity of POAG was 1.81.

   • Both the presence and number of first-degree relatives with glaucoma contribute to POAG (OR 7.28 and 6.61, respectively). However, unlike PAC, only siblings and offspring with glaucoma contributed to the odds of having POAG (OR 8.99 and OR 19.23, respectively)—not parents.2764

Researchers at two centers in the Netherlands conducted a study to determine the 10-year incidence of glaucomatous visual field loss (GVFL) in the offspring of individuals with glaucoma. Participants were 94 offspring of glaucoma cases from the Rotterdam Study who were examined in 1994 and 2000 and followed-up in 2007-2008.

After a mean follow-up duration of 11.5 years (SD 2.0; r: 7.4 to 13.7 years), four participants with GVFL were identified (mean age at follow-up, 59 years): three incident cases and a single subject who already had GVFL at baseline, for a prevalence of 4.3 percent in this group—significantly higher than the prevalence of 0.7 percent in the general population of similar age. The incidence rate and 10-year risk of GVFL were 2.8/1000 person-years and 2.8 percent respectively. Factors associated with GVFL were: high myopia (p=0.003); IOP (p=0.032); IOP treatment (p=0.004) and linear cup-to-disc ratio (p=0.009).

The authors note that although being an offspring of a glaucoma case does appear to be a risk factor for developing visual field loss, a 10-year risk of only 2.8 percent suggests that examinations of offspring can be widely spaced in time.3979

 

The Cornea Connection

Researchers at two American universities, along with the Ocular Hypertension Treatment Study Group and European Glaucoma Pre-vention Study Group, conducted a study to determine whether adjusting baseline IOP for corneal thickness actually does increase the accuracy of the prediction model for the development of POAG from the Ocular Hypertension Treatment Study and the European Glaucoma Prevention Trial. (This study was supported by multiple non-commercial grants along with grants from Merck, Pfizer and RPB.)

The authors ran the same Cox proportional hazards model using each of five published correction algorithms, changing only whether or not IOP was adjusted for CCT. They found that, in the models that included IOP adjusted for CCT, the C-indices ranged from 0.763 to 0.770—no better than the original prediction model (0.774). When CCT was added to models in which IOP was adjusted for CCT, CCT still made a statistically significant improvement.

The authors conclude that calculating individual risk for the development of POAG is simpler and just as accurate using IOP and CCT as measured, as when applying an algorithm to correct IOP for CCT.2708

Researchers at the University Eye Hospital Freiburg, in Freiburg, Germany, conducted a study to determine whether corneal hysteresis, as measured by the Ocular Response Analyzer (Reichert), really describes an independent biomechanical property of the cornea or is itself dependent on IOP. Nine donor corneas were clamped into an artificial anterior chamber and corneal hysteresis measurements were obtained while manually adjusting the IOP.

The data showed a highly significant association between corneal hysteresis and IOP: CH decreased as IOP increased. The authors note that this data suggest that the CH as measured by the ORA is not an independent biomechanical property of the cornea. Either CH itself is dependent on IOP, or the ORA is incapable of measuring CH accurately at higher IOP levels.5553

Researchers at three centers in New York State prospectively investigated the correlation between CCT and corneal hysteresis and their relationship to the rate of visual field progression. One hundred fifty-three eyes of 153 glaucoma patients with at least five SITA Standard 24-2 VF tests showing baseline glaucomatous damage were included, with a mean follow-up time of 5.3 years.

The data showed:

   • Progressing eyes had thinner CCT (525.0 ±34.2 vs. 542.3 ±38.5 µm, p=0.04) and lower CH (7.5 ±1.4 vs. 9.0 ±1.8 mmHg, p<0.01) compared to nonprogressing eyes.

   • CH and CCT correlated significantly (Pearson's r=0.33, p<0.01).

   • Older age (OR=1.04, p=0.04), CH (OR=0.63, p<0.01), and mean follow-up IOP (OR=1.24, p<0.01) were significantly associated with VF progression.

The authors note that because CH is associated with increased risk of VF progression and may describe corneal properties more completely than thickness alone, it may be a better parameter to assess the risk of progression.4025

 

New Diagnostic Tools

Researchers at Gulhane Military Medical Academy in Ankara, Turkey, conducted a prospective study to determine whether glaucoma affects the functional recovery of the macula under photostress testing conditions. They enrolled two groups: 20 adults (eight men and 12 women, mean age 69.9 ±7.32 years) who had primary open-angle glaucoma with established visual field defects; and 20 age- and sex- matched healthy subjects (mean age 68.05 ±7.28 years) as a control.

Both groups underwent visual field testing using the Humphrey Perimeter (Central 24-2 SITA-FAST). While in front of the perimeter, they underwent photostress testing with a standard 60-watt white light bulb 50 cm from the eye. In addition, Stratus OCT was used to determine thickness of the central macula and retinal nerve fiber layer.

The data showed that all eyes in the glaucoma group had lower foveolar threshold values (p<0.001) and longer recovery times (p<0.001) than the controls. No significant correlation existed between CMT and recovery time, but there was a moderate correlation between RNFL and recovery time (p<0.01). The authors note that photostress testing may be useable as a functional biomarker of glaucoma.5492

In hopes of finding a screening test for glaucomatous optic nerve damage that's quick, inexpensive and easily performed by trained allied health-care workers, researchers at Stanford University School of Medicine conducted a study to see whether testing relative afferent pupillary defect (RAPD) using a swinging flashlight could be an effective tool for glaucoma screening. Non-ophthalmic personnel masked with regard to disease presence examined 107 subjects recruited from a mixed population of glaucomatous and non-glaucomatous patients. All subjects underwent a swinging flashlight test with, when necessary, the aid of neutral density filters, to determine whether or not a RAPD was present. Medical charts were subsequently reviewed to grade disease stage.

Statistical analysis found an OR of 9.71 for glaucoma if RAPD was present, with a sensitivity of 66.7 percent and a specificity of 82.9 percent. A planned subanalysis of phakic study subjects revealed an OR of 17.05 for glaucomatous disease if RAPD was present, with a sensitivity of 68.8 percent and a specificity of 88.6 percent. Sensitivity, specificity and predictive value was greater for phakic subjects than for the overall group, which included those who had previously undergone cataract surgery.183

 

Dr. Netland is professor and chairman of the department of ophthalmology at the University of Virginia; Dr. Singh is professor of ophthalmology and director of the Glaucoma Service at Stanford University School of Medicine.