JOURNAL CLUB

RECENT NOTEWORTHY STUDIES TO STIMULATE DISCUSSION AND DEBATE

■ ERG in diabetic retinopathy. It's always been clear that diabetes can adversely affect vision, but at what point do retinal defects translate into functional changes? A team of retinal physicians collaborating between Hong Kong and Australia used electroretinography to answer that question. Their findings were published online on April 12 by Graefe's Archive of Clinical and Experimental Ophthalmology.

The authors enrolled 38 diabetic patients, nine of whom were free from diabetic retinopathy (Figure 1) and the remainder of whom had mild to moderate DR. These patients and 14 age-matched controls were submitted to global flash multifocal electroretinography at high and low contrast levels, with the Z scores of the responses compared with regard to the different types of retinal defects detected.

Figure 1. Electroretinography was shown to be effective in detecting functional changes due to diabetic retinopathy (seen above).

Local reductions in both direct component and indirect component amplitudes were found in the diabetic patients, regardless of whether they had developed DR. Furthermore, as DR developed and increased in severity, there was a correlating deterioration of amplitude in both components, demonstrating the efficacy of ERG in screening for DR, apparently even before visible signs of DR appear. ERG may emerge soon as a regular screening tool for the diagnosis and follow-up of DR.

■ Retinal breaks and vitrectomy. The May 2012 issue of the American Journal of Ophthalmology features a study from New Zealand on the frequency of peripheral iatrogenic retinal breaks in eyes undergoing small-gauge pars plana vitrectomy. To determine this frequency, the authors undertook a prospective, single-center, noncomparative, interventional case series of 184 patients undergoing PPV.

The enrolled patients, who had a mean age of 65.6 years old, underwent either 23- or 25-gauge PPV. The reasons for surgery included epiretinal membrane, macular hole, vitreous hemorrhage, tractional retinal detachment due to proliferative DR, vitreomacular traction and floaters. The authors found that of a total of 29 retinal breaks identified in 184 patients, more than a third of these breaks occurred in patients submitted to surgery for vitreous hemorrhages, with the next largest cohorts consisting of patients with tractional retinal detachments due to proliferative DR and epiretinal membranes.

When determining the number of these 29 breaks due to sclerotomy, only six still qualified, with half of them occurring in patients with vitreous hemorrhages. The authors found that entry-site breaks were not related to instrument gauge, although there was a statistically insignificantly greater likelihood of breaks in any part of the retina occurring with 23-gauge instruments. Only one rhegmatogenous retinal detachment occurred over the three-month follow-up interval.

The study authors are quick to clarify that the study has several limitations, not the least of which are its relatively small size and lack of randomization. Nevertheless, its prospective nature is important in providing greater evidence for a lack of risk for retinal breaks due to small-gauge vitrectomy.

■ Retinal surgery and keratoprostheses. With increasing use of keratoprostheses by anterior-segment surgeons, the issue of how to undertake vitreoretinal surgery in the presence of these devices has become more urgent. In the April 2012 issue of Archives of Ophthalmology, doctors from several New York medical centers collaborated to address these issues.

Twenty-three small-gauge vitrectomies, performed in 13 patients, were studied in this retrospective review. All of the patients received a permanent Boston type 1 keratoprosthesis (KPro), either before vitreoretinal surgery or simultaneously with the posterior-segment procedure. Among the indications for posterior-segment surgery were eight cases of retro-KPro membranes, as well as retained lens fragments, endophthalmitis and retinal detachments, among other problems.

Of the eight cases of retro-KPro membranes, seven of them occurred in eyes in which vitrectomy was not performed simultaneously. There were no other serious complications in this case series, and visual acuity improved in most of the cases. The authors conclude that retro-KPro can best be avoided by performing keratoprosthesis implementation and vitrectomy at the same time.

■ Lucentis cost-efficacy in DME. Using data from the RESTORE phase 3 trial of Lucentis monotherapy vs combination therapy with laser, European, Australian and US ophthalmologists collaborated on a study to evaluate the cost-effectiveness of Lucentis alone and in combination in patients with diabetic macular edema (Figure 2). They report their findings in the May 2012 issue of the British Journal of Ophthalmology.

Figure 2. Diffuse diabetic macular edema (DDME); macular view of ultra-widefield fluorescein angiogram (a) and SDOCT (b).

Studying the 345 patients enrolled in RESTORE, the authors designed a Markov model to simulate the long-term outcomes and costs of treating DME in one eye. The outcomes they measured included quality-adjusted life years (QALYs) and the costs of treatment, disease monitoring, and visual impairment and vision loss.

Use of Lucentis as monotherapy resulted in a mean increase of 0.17 QALY at a costs of £4,191(approximately $6,750) or £24,028 per QALY (almost $40,000), resulting in a 64% probability of being cost-effective at a threshold of £30,000 (almost $50,000) per QALY. In combination with laser, Lucentis was actually less cost-effective, costing £4,695 per QALY (~$7,575) and only a 42% probability of being cost-efficient at a threshold of £30,000.

While the authors decline to conclude that the cost-effectiveness of combination therapy with Lucentis and laser is less than that of Lucentis monotherapy, they nevertheless conclude that Lucentis monotherapy is cost-effective. It should be noted that several of the authors of this study are associated with Novartis, which distributes Lucentis.

■ Submacular hemorrhage in AMD. Recurrent submacular hemorrhage is a major complication of wet AMD, with thicker hemorrhages having poor prognoses because of the scarring they cause. In patients in whom such hemorrhages recur, prognoses are even worse. In the April 2012 issue of Retina, authors from Seoul attempt to identify the risk factors and long-term outcomes associated with recurrent submacular hemorrhages.

The medical records of wet AMD patients with submacular hemorrhaging were reviewed by the authors. All of the records included follow-up data of at least two years. Any newly developed hemorrhage larger than one disc area, after near-complete resolution of the initial hemorrhage, was defined as a recurrence.

The patients were divided into two groups, depending on whether they had recurrent submacular hemorrhages. Twenty-four of the 47 patients included had such recurrences. The average time to recurrence was 21.4 ±9.2 months. Importantly, half of the patients in the group having recurrences also had polypeptide choroidal vasculopathy (which is more common among Asian subjects), while only 13% of the patients without recurrence had PCV.

A total of 70.8% of the recurrence patients and 95.7% of the other patients were treated with anti-VEGF agents. There was no statistically significant difference in final visual acuity between the two groups.

The authors of the study concluded that the rate of recurrence of submacular hemorrhage in the cohort they studied was 51.1%, with the presence of PCV strongly associated with the risk of recurrence. Anti-VEGF agents reduced the risk of recurrence, they reported. They urge that further, prospective trials be undertaken to test anti-VEGF agents in such patients.

■ FUSION data. On April 15, Graefe's published online the results of a one-year study of the FUSION regimen used in patients with wet AMD who had relatively good baseline visual acuity. The FUSION regimen consists of a loading phase of two or three injections of Lucentis, depending on the presence of CNV; one injection upon the disappearance of exudation; and subsequent administration of two injections at two-month intervals, followed by injections every three months.

The study enrolled 17 eyes of 17 patients and found that at three months, mean BCVA had improved by 2.3 letters. At six months, the mean change in BCVA was +4.2 letters. At 12 months, the mean change in BCVA was +5.6 letters. The mean number of injections over one year was 6.9. Only one patient dropped out of the study, and one other patient had an RPE tear, but there were no other adverse events.

The authors hope that their study provides more evidence for an approach to treatment with Lucentis that would dictate retreatment before the reappearance of retinal fluid. This approach could be particularly effective in patients with decent baseline BCVA. RP