JOURNAL CLUB

RECENT NOTEWORTHY STUDIES TO STIMULATE DISCUSSION AND DEBATE

ANDREW E. MATHIS, PhD, MEDICAL EDITOR

■ Making the grade. A pair of articles in the June 2010 issue of the British Journal of Ophthalmology address issues with the grading of diabetic retinopathy. The first, a study prepared by members of the Scottish Diabetic Retinopathy Clinical Research Network, assessed the role of hemorrhage and exudate detection in automated grading.

Taking images from more than 8,500 patients from three grading centers in Scotland, the study authors then reference-graded all the images. Automated disease/no disease assess ments were then made based on detection of microaneurysms, hemorrhage and exudates. The authors conclude with the suggestion that automated disease/no disease grading could be a powerful adjunct to current grading systems in use, particularly as screening expands to less developed countries.

The other BJO study, also from the Scottish Diabetic Retinopathy Clinical Research Network, compared the costs and consequences of automated vs manual grading for the detection of diabetic retinopathy. Using the same images from the first study, the authors here compared a newer algorithm for grading with the one currently in use.

The new algorithm was found to be more cost-effective than the current one, saving £215.75 (approximately $325) per additional appropriate screening outcome. Ultimately this would save as much as £3,834 per case missed in costs of care, totalling perhaps as much as £267,115 over an extrapolated 20-year period in the model studied.

Not surprisingly, the authors recommend that this newer automated grading system be introduced in Scotland.

■ More diabetic retinopathy research. Elsewhere in Europe, a team of retinal physicians has examined the purported link between proliferative diabetic retinopathy in one eye and the presumably higher risk of neovascularization in the other eye. The Icelandic authors report their results in the June 2010 issue of Acta Ophthalmologica.

Examining a database of over 1,500 patients with diabetes, 76 patients with PDR were identified and studied. Twentyeight had bilateral disease at the beginning of the study period, with another 28 developing bilateral disease within five years of diagnosis in the first eye. Mean duration of diabetes before development of PDR was 19 years for type 1 diabetes and 14 years for type 2.

Among the risk factors identified in the study were poor hemoglobin control among all patients with PDR, as well as a higher rate of smoking in the group with type 2 diabetes. The authors conclude that the risk of developing retinopathy in the fellow eye when it exists in one eye is indeed genuine, but they concede that their study is too small to be significantly powered.

■ Crystalline maculopathy in triamcinolone-treated eyes. Along with the already well-established risks of increased intraocular pressure and development of cataracts, another potential side effect of intravitreal injections of triamcinolone acetonide may be crystalline maculopathy. According to a study in the June 2010 issue of Archives of Ophthalmology, 21 eyes in 13 patients, all of whom were treated for macular edema with triamcinolone in Los Angeles and New York over a nine-year period, developed superficial, refractile, white or yellow-green crystals in the posterior pole.

The upshot of the study is that these cases of crystalline maculopathy were benign and did not lead to visual deficit. The authors posit that the crystals result from larger-thanaverage triamcinolone particles in injection suspensions resisting washout from patients' systems over time. The association of crystalline maculopathy with triamcinolone use varied among three formulations of the drug; two showed a correlation while one did not.

The authors recommend that retinal physicians add this particular type of maculopathy to differential diagnosis when examining patients previously treated with triamcinolone.

■ Vitrectomy in advanced AMD. When should vitrectomy be the choice for treating vitreous hemorrhage in advanced cases of wet AMD? Two articles in the June 2010 issue of Retina examine the issue. The first article, a Japanese study, consists of a retrospective case series of 60 eyes that underwent pars plana vitrectomy, having received no other treatment.

After vitrectomy, 40 eyes either regressed in their hemorrhage or had the bleeding disappear entirely. Both the incidence of bleeding and postoperative visual acuity were better compared to the six-month preoperative levels to statistically significant extents. A key finding was that choroidal neovascularization improved more in eyes without posterior vitreous detachment, leading the authors to conclude that pars plana vitrectomy should be a frontline treatment in cases of AMD where there is strong vitreomacular traction.

The second study, conducted in Korea, evaluated vitrectomy when performed on 24 eyes of 23 patients, half of whom had exudative AMD and the other half of whom had polypoidal choroidal vasculopathy. Improvement in visual acuity was seen at higher levels in the PCV group.

Regardless of outcomes, the study authors underscore the importance of screening for disciform scarring before performing vitrectomy on patients with either AMD or PCV. However, they do stress that, in eyes with PCV with vitreous hemorrhage, vitrectomy may be an effective intervention.

■ Eyedrops for DME. A Japanese study published in the June 2010 issue of Graefe's Archive for Clinical and Experimental Ophthalmology examined a steroid eyedrop vs subconjunctival triamcinolone injection for the reduction of refractory diabetic macular edema, with the eyedrop comparing favorably.

All 22 patients enrolled in this trial had previously undergone pars plana vitrectomy, with many also having been treated with macular or panretinal photocoagulation and/or intravitreal or sub-Tenon's injection of triamcinolone. No patients were treatment-naïve. After three months of treatment with difluprednate ophthalmic emulsion (Durezol, Alcon), mean visual acuity in the eyedrop-treated group had not improved appreciably, but mean retinal thickness had decreased from 500.6 ±207.7 μm at baseline to 341.2 ±194.8 μm, compared to comparable drop in central retinal thickness in the triamcinolone-treated group, which did experience an improvement in visual acuity. The rate of effective improvement in retinal thickness did not differ between groups.

The authors conclude that the eyedrops are a safe and effective treatment, particularly because they can be applied following vitrectomy, they do not require an invasive procedure, and they do not incur serious adverse events.

■ High IOP with anti-VEGF. Evaluating claims of higher IOP in patients treated with either ranibizumab or bevacziumab, doctors from the University of Colorado have published an article in the June 2010 issue of Retina looking at the possibility that high–molecular-weight aggregates in the packaged products are responsible.

The authors used enzyme-linked immunosorbent assay, size-exclusion chromatography and polyacrylamide gel electrophoresis to analyze repackaged bevacizumab obtained from three different compounding pharmacies and determine immunoglobulin G (IgG) levels. All the syringes examined contained statistically similar IgG levels, although one pharmacy's solution had significantly more functional IgG. The two other pharmacies' compounds had 10 times more micron-sized particulates.

The authors conclude that significant differences in IgG levels and particulate size are related and have an impact on the likelihood of raised IOP in patients receiving bevacizumab injections. Given the need to repackage bevacizumab for ophthalmic use, whereas ranibizumab is already formulated for use in the eye, ranibizumab was not studied here.

■ Anti-VEGF for RAP. Doctors from USC and UCLA have concluded that anti-VEGF therapy can be a useful option in treating patients with retinal angiomatous proliferation, reporting their findings in the June 2010 issue of the British Journal of Ophthalmology. Twenty eyes in 15 patients with RAP lesions were treated with either ranibizumab, bevacizumab or both and followed up at one, three, and six months.

At all follow-up visits, mean visual acuity improved, from 20/72 at baseline to 20/52 at one month, 20/45 at three months, and 20/56 at six months. Then, at the twoyear mark, improvement of visual acuity was maintained in 62.5% of the patients treated.

The study adds to the growing body of research suggesting that anti-VEGF drugs can be helpful in treating patients with RAP and other types of advanced neovascular lesions.

■ Early AMD risk factors. The Beaver Dam Offspring Study has provided more information on risk factors in AMD, with the latest results appearing in the June 2010 issue of Archives of Ophthalmology.

In the cohort of 2,810 patients who were evaluated for the presence of drusen and lesions typically seen in AMD, 3.4% were identified as having early AMD, with the rate being 9.8% in patients 65 or older. (The study includes patients as young as 21.) Age, male gender and smoking were all identified as risk factors, as well as higher serum HDL cholesterol levels and hearing impairment.

Conversely, no relationship was seen between the development of early AMD and blood pressure, body mass, physical activity, alcohol use, white blood cell counts and serum total cholesterol.

While this study confirms the previous finding that early AMD is very rare in patients below 55 years old (in this study, 2.4% of patients in the 21-to-34-year-old group had early AMD), other assumptions, such as that atherosclerosis of the choroid and lipid deposition in Bruch's membrane lead to early AMD, were not borne out. The authors do point out that the demographics of the study may have been skewed to lower socioeconomic and educational levels, which may account for some of the factors that were evaluated here. Nevertheless, prevention remains the key in early AMD. RP