Milton C. Chew

Measuring the precise area of peripheral retinal non-perfusion using ultra-widefield imaging and its correlation with the ischaemic index

Objective To determine the calculated, anatomically correct, area of retinal non-perfusion and total area of visible retina on ultra-widefield fluorescein angiography (UWF FA) in retinal vein occlusion (RVO) and to compare the corrected measures of non-perfusion with the ischaemic index. Methods Uncorrected UWF FA images from 32 patients with RVO were graded manually for capillary non-perfusion, which was calculated as a percentage of the total visible retina (uncorrected ischaemic index). The annotated images were converted using novel stereographic projection software to calculate precise areas of non-perfusion in mm2, which was compared as a percentage of the total area of visible retina (‘corrected non-perfusion percentage’) with the ischaemic index. Results The precise areas of peripheral non-perfusion ranged from 0 mm2 to 365.4 mm2 (mean 95.1 mm2), while the mean total visible retinal area was 697.0 mm2. The mean corrected non-perfusion percentage was similar to the uncorrected ischaemic index (13.5% vs 14.8%, p=0.239). The corrected non-perfusion percentage correlated with uncorrected ischaemic index (R=0.978, p<0.001), but the difference in non-perfusion percentage between corrected and uncorrected metrics was as high as 14.8%. Conclusions Using stereographic projection software, lesion areas on UWF images can be calculated in anatomically correct physical units (mm2). Eyes with RVO show large areas of peripheral retinal non-perfusion.

Comparison of foveal-sparing with foveal-involving photodynamic therapy for myopic choroidal neovascularization

PurposeTo evaluate the visual outcomes of choroidal neovascularization (CNV) secondary to pathological myopia in eyes treated with photodynamic therapy (PDT), and to determine the effect of lesion location and foveal involvement on visual prognosis.MethodsInterventional case series of 24 consecutive patients with myopic CNV treated with PDT. The main outcome measure was final LogMAR visual acuity (VA).ResultsOf 24 eyes, the CNV lesion was subfoveal in 11 and extrafoveal in 13. Overall, the mean LogMAR VA at 24 months was 0.72. Extrafoveal CNV lesions achieved significantly better final VA compared with subfoveal CNV (LogMAR 0.45 vs 1.05, P=0.012). Eyes with extrafoveal CNV lesions were subdivided into foveal-sparing PDT (where the PDT laser spot did not involve the foveal center) and foveal-involved PDT (where the PDT laser covered the fovea). At all time points, the group with foveal-sparing PDT had significantly better VA compared with the foveal-involved group. The final LogMAR VA for the foveal-sparing PDT group was 0.26 compared with 1.00 for the foveal-involved PDT group (P=0.003). At 24 months, 77.8% of foveal-sparing PDT cases achieved VA of ≥20/40, compared with 25% of foveal-involved PDT cases and 9.1% of subfoveal CNV lesions (P=0.006).ConclusionFor patients with myopic CNV, foveal-sparing PDT results in significantly better long-term visual outcomes compared with those with foveal-involved PDT. Foveal-sparing PDT may be of value for treatment of myopic CNV patients who are not suitable for treatment with anti-vascular endothelial growth factor injections.

Diurnal variation of retinal thickness measured by optical coherence tomography in normal adults.

We read with interest the article by Jo et al. in the August 2011 issue describing the diurnal variation of retinal thickness in normal subjects measured by time-domain (TD) and spectraldomain (SD) optical coherence tomography (OCT). The authors concluded that the diurnal variation in the retinal thickness observed with TD-OCT was caused by limitations in the repeatability of the OCT device rather than by actual variation of retinal thickness. Although this was a well-designed study, one limitation was that the retinal thickness was assessed at only two time points, which provides limited information on the pattern of diurnal change. A difference between two time points is less compelling than if the retinal thickness measurements differ across several additional points during the day. We would like to highlight two other studies in normal subjects that provide additional evidence to support the findings of Jo et al. A significant strength of both of the additional studies is that the retinal thickness was measured at five time points during a single day, thus providing more comprehensive and robust data for the analysis of diurnal variation compared with the two time points that were used in the present study. In addition, in both studies, the examination sequence (five OCT scans during the day) was repeated on a separate day to confirm the consistency of the results. In a study of 12 normal individuals, sequential SD-OCT scans were performed with the Spectralis OCT (Heidelberg Engineering, Heidelberg, Germany). The authors found no significant variation in the macular thickness over the five time points during the day, with a maximum amplitude (the difference between the maximum and minimum retinal thickness) of 0.9 m. This result is comparable to the difference of 0.3 m reported by Jo et al. for central macular thickness measured with the Cirrus OCT (Carl Zeiss Meditec, Dublin, CA). In addition, this study used the tracking function of the Spectralis OCT to reduce the chance of a small displacement of the foveal B-scan position between sequential OCT scans and to ensure that the same point on the retina was measured for all OCT scans. We feel that this is a useful and important advantage, conferred by the tracking function of the Spectralis OCT. In another study, Chakraborty et al. 3 used an optical biometer instead of OCT to measure retinal thickness. They also reported no significant diurnal variation in retinal thickness (mean amplitude, 8 m) over the 2 days of examination. Another interesting point that merits discussion is that, in Jo et al., in the sectors that showed a significant difference in retinal thickness on TD-OCT, the mean difference ranged from 2.5 to 3.5 m. Although this difference may be statistically significant, it is unlikely to be of clinical relevance. In addition, the coefficient of repeatability of Stratus OCT retinal thickness in normal individuals has been reported to be approximately 17 m, which is much greater than the difference between the two time points in this study. Jo et al. also cited earlier studies that have reported diurnal variation in retinal thickness in patients with macular diseases. In these studies, various versions of TD-OCT were used to perform the OCT scans. Although it is certainly possible that the diurnal variation in retinal thickness may be due to the underlying disease process, it is also possible that at least part of it is due to motion artifact resulting from poor fixation as a result of the underlying disease process. It would be interesting to know whether a similar degree of diurnal change in retinal thickness in diseased eyes would be observed with tracking-capable SD-OCT devices. In summary, in confirming this finding using SD-OCT with tracking, we agree with Jo et al. that there does not appear to be significant diurnal variation in retinal thickness in normal individuals. Colin S. H. Tan Wei Kiong Ngo Milton C. Chew Kelvin Z. Li Louis W. Lim Srinivas R. Sadda

Advances in retinal imaging for diabetic retinopathy and diabetic macular edema

Diabetic retinopathy and diabetic macular edema (DME) are leading causes of blindness throughout the world, and cause significant visual morbidity. Ocular imaging has played a significant role in the management of diabetic eye disease, and the advent of advanced imaging modalities will be of great value as our understanding of diabetic eye diseases increase, and the management options become increasingly varied and complex. Color fundus photography has established roles in screening for diabetic eye disease, early detection of progression, and monitoring of treatment response. Fluorescein angiography (FA) detects areas of capillary nonperfusion, as well as leakage from both microaneurysms and neovascularization. Recent advances in retinal imaging modalities complement traditional fundus photography and provide invaluable new information for clinicians. Ultra-widefield imaging, which can be used to produce both color fundus photographs and FAs, now allows unprecedented views of the posterior pole. The pathologies that are detected in the periphery of the retina have the potential to change the grading of disease severity, and may be of prognostic significance to disease progression. Studies have shown that peripheral ischemia may be related to the presence and severity of DME. Optical coherence tomography (OCT) provides structural detail of the retina, and the quantitative and qualitative features are useful in the monitoring of diabetic eye disease. A relatively recent innovation, OCT angiography, produces images of the fine blood vessels at the macula and optic disc, without the need for contrast agents. This paper will review the roles of each of these imaging modalities for diabetic eye disease.

Treatment options for myopic CNV - Is photodynamic therapy still relevant?

Dear Sir, We read with interest the article by Manayeth et al.[1] describing the use of low-fluence photodynamic therapy (PDT) to successfully treat a patient who developed myopic choroidal neovascularization (CNV) following laser in situ Keratomileusis (LASIK), as well as the comment by Gopal[2] on the efficacy of anti-vascular endothelial growth factor (anti-VEGF) agents compared to PDT in the treatment of CNV. The Ranibizumab and PDT [verteporfin] evaluation in myopic choroidal neovascularization (RADIANCE) study,[3] a randomized controlled trial comparing ranibizumab against verteporfin PDT for the treatment of myopic CNV, reported that ranibizumab treatment provided superior visual acuity (VA) gains compared to PDT. In addition, among the non-randomized studies published in the literature, patients with myopic CNV treated using anti-VEGF agents were reported to have better mean VA compared to patients treated with PDT.[4] However, we would like to highlight that PDT may result in good visual outcomes in carefully selected patients, especially those with extrafoveal CNV lesions where the laser spot can be adjusted to spare the fovea. In a study of 24 eyes with myopic CNV,[5] we found that among patients who were treated with foveal-sparing PDT, 77.8% achieved VA of ≥ 20/40, with a mean final logMAR VA of 0.26. An additional factor influencing the outcome is size of the myopic CNV lesion. Tan et al.[4] reported that myopic CNV lesions with a greatest linear diameter (GLD) of ≤ 1000 μm had better outcomes compared to those with larger GLD. While we acknowledge the difficulty of making direct comparisons of results from different studies, it is interesting to note that the visual outcomes in our study were comparable to, or in some cases better than, those reported from studies using anti-VEGF agents.[5] Anti-VEGF agents are associated with systemic risks such as cerebrovascular accidents and other arterial thromboembolic events, especially for patients with pre-existing disease.[5] In addition, intravitreal injections carry the risk of infectious endophthalmitis.[5] Some patients are unwilling to accept either the systemic or ocular risks associated with anti-VEGF agents, and for them, PDT may be a valuable modality of treatment. In conclusion, ophthalmologists may wish to consider PDT in cases where the fovea can be spared, and where anti-VEGF agents are unsuitable or unacceptable to patients. It has been shown that extrafoveal CNV lesions occur in between 18.5 and 32% of myopic CNV patients.[5] Therefore, we believe that PDT may still have a useful role in the management of myopic CNV, if patients are carefully selected.

Factors affecting visual outcome of myopic choroidal neovascularization treated with verteporfin photodynamic therapy.

AIM To evaluate the visual outcomes of choroidal neovascularization (CNV) secondary to pathological myopia and the impact of novel risk factors affecting the final visual outcome. METHODS Interventional case series of 18 consecutive patients with pathological myopia treated with photodynamic therapy (PDT). Inclusion criteria were spherical equivalent -6D or worse or features of pathological myopia on retinal examination. The main outcome measure was final best-corrected visual acuity (BCVA). RESULTS Of 18 eyes, 13 (72.2%) avoided moderate visual loss (≥3 lines of LogMAR BCVA) and 5 eyes (27.8%) improved by at least 1 line after 1 year. Patients with LogMAR BCVA ≤0.3 (Snellen equivalent 20/40) at one year were younger than those with BCVA >0.3 (mean age 39.0 vs 61.6 years, P=0.001). A higher proportion of eyes with greatest linear dimension (GLD) of ≤1000µm avoided moderate visual loss (100% vs 50%, P=0.026). Among patients who were treated within 2 weeks of visual symptoms, 88.9% avoided the loss of 3 or more lines compared to 55.6% for those who presented later. The mean improvement in LogMAR BCVA of those with GLD ≤1000µm was +0.12 compared to a loss of 0.55 LogMAR units for those with GLD >1000µm (P=0.02). Visual outcomes were not associated with gender or refractive error. CONCLUSION Good visual outcome in myopic CNV is associated with younger age, smaller lesion size and earlier initiation of treatment. These factors are relevant for ophthalmologists considering treatment options for myopic CNV.

Effect of Photocoagulation of Ischemic Areas to Prevent Recurrence of Diabetic Macular Edema

We read with interest the article by Takamura et al. describing the outcomes of photocoagulation of ischemic areas to prevent recurrence of diabetic macular edema (DME) among patients treated with focal/grid laser photocoagulation and intravitreal bevacizumab injection. The authors described a sustained improvement in central retinal thickness (CRT) in the group that received targeted retinal photocoagulation (TRP) compared to the group that did not receive TRP. In their article, the authors mentioned that the therapeutic effects of anti-VEGF agents seem to be transient and suggested that the results of their study ‘‘may imply that multiple injections are still needed to maintain the therapeutic effect of an anti-VEGF drug for a longer period even if TRP was successfully performed.’’ While we agree with the authors that patients with DME often require multiple injections of anti-VEGF to maintain the gain in visual acuity and reduction in CRT, it is important to note that the number of injections required typically decreases after the first year. A study by the Diabetic Retinopathy Clinical Research Network reported that patients with DME received a median of eight to nine anti-VEGF injections within the first year of treatment. However, the number of treatments decreased to two to three in the second year and one to two in the third year. This is an important consideration when counseling patients on the expected course of the disease. The authors mentioned that it would have been ideal to assess the amount of retinal nonperfusion in the peripheral regions of the retina using ultrawide field (UWF) imaging systems, and correctly pointed out that UWF imaging can capture up to 2008 of the retina in a single image. This is a considerably larger area than the 758 covered by the seven standard fields (7SF), and it has been demonstrated that pathology that is missed on 7SF may be detected using UWF imaging. One possible explanation for the requirement for additional treatment with anti-VEGF injections despite successful TRP was that additional untreated areas of peripheral retinal nonperfusion, which was not seen within the field of view of 7SF imaging, continued to produce VEGF, with resultant recurrence of macular edema. Another potential explanation is progression or enlargement of areas of retinal nonperfusion. Studies have reported changes in areas of retinal nonperfusion among patients with retinal vein occlusion, and it is possible that this occurs in diabetic retinopathy as well. Regions of the retina that appeared to be perfused at the time of initial fluorescein angiography (FA) may be watershed areas, where perfusion is variable during the course of the disease. We are curious whether additional FA was performed during the course of the disease, and whether these showed changes in areas of retinal nonperfusion. In summary, we congratulate the authors on an interesting study, and fully agree with them that assessment of ischemic areas using UWF FA is essential in the management of patients with retinal vascular diseases.

Prophylactic intracameral antibiotic use during cataract surgery

In a recent article, Kam et al. described the frequency of intracameral antibiotic usage following cataract surgery among cataract surgeons in Victoria, Australia. The adoption rate of intracameral antibiotics among surgeons in Victoria are indeed impressive, and the authors reported a corresponding decrease in the rate of endophthalmitis seen at the Royal Victorian Eye and Ear Hospital during the period following the increased adoption of intracameral antibiotics. From the authors’ results, intracameral cefazolin was used by 72.9% of respondents, either alone or in combination with subconjunctival cefazolin. In contrast, from the chart in figure 1 of their article, subconjunctival gentamicin (either given in combination with cefazolin or alone) appears to be used by less than 10% of cataract surgeons, with a small number giving only subconjunctival gentamicin. We are curious whether the survey respondents were asked for the rationale behind their choice of antibiotics. In most reports, Gram-positive organisms were the most common causative organisms of postoperative endophthalmitis, and many centres use cephalosporins because of the coverage of Gram-positive organisms (63–70%). At our centre, both intracameral cefazolin and subconjunctival gentamicin are used at the end of surgery in order to provide prophylaxis against both Gram-positive and Gram-negative organisms. This is relevant because we found that among our patients with endophthalmitis, 25% of culture-positive patients had Gram-negative organisms such as pseudomonas sp. The rate of culture of Gram-negative organisms from endophthalmitis reported in some Asian populations is higher than in Caucasian populations (19–25%), and we are curious about the types of organisms cultured from the patients as seen at Royal Victorian Eye and Ear Hospital. If the rates of Gram-negative organisms cultured are relatively high, then the use of an antibiotic, which provides greater coverage of Gram-negative organism such as gentamicin, may be a relevant consideration. In their manuscript, Kam et al. commented on the lack of a treatment arm that utilized prophylactic subconjunctival cefuroxime in the study by the European Society of Cataract and Refractive Surgeons. In view of the weight of current evidence for the efficacy of intracameral antibiotics, it is unlikely that such a comparison will be performed in a future randomized controlled trial. The authors mentioned non-randomized studies demonstrating a decrease in the rate of endophthalmitis following the use of intracameral cefuroxime. We would like to highlight that similar trends were observed when cefazolin was used. In a review of 50 177 cases of cataract surgery, the rates of endophthalmitis was 0.064% when subconjunctival cefazolin was used, and decreased significantly to 0.01% following a switch to intracameral cefazolin (multivariate odds ratio 13.6, P < 0.0001). Like the authors, we also recognize that this study compared two cohorts over different time periods, and the results may have been influenced by other factors such as improvements in surgical technique or equipment. However, in the absence of level 1 evidence from a randomized controlled trial, we feel that these findings are important for ophthalmologists to consider when assessing the utility of prophylactic antibiotics. In summary, we fully agree with the authors that it is essential for evidence of the efficacy of prophylactic intracameral antibiotics in reducing the rates of endophthalmitis to be disseminated to cataract surgeons, and applaud their important contributions to this effort.

Letter to the editor: Forty-two-month outcome of intravitreal bevacizumab in myopic choroidal neovascularization.

Dear Editor, We read with interest the article by Traversi et. al. [1] reporting the long-term visual outcomes of treatment-naïve myopic choroidal neovascularization (mCNV) treated with intravitreal bevcizumab injections. Their results are impressive, demonstrating stable gains in visual acuity lasting up to 42 months following treatment. The authors mentioned that photodynamic therapy (PDT) has failed to demonstrate statistically significant benefits after 2 years based on the Verteporfin in Photodynamic Therapy (VIP) study [2]. However, we would like to highlight that the VIP study evaluated only patients with subfoveal mCNV.We believe that PDTstill has a role in the management of extrafoveal mCNV, which may occur in between 18.5% to 32% of myopic CNV patients [3]. Tan et. al. [3] recently demonstrated that if the PDT laser spot does not involve the center of the fovea in extrafoveal mCNV, the visual outcomes are comparable to longterm studies using anti-VEGF therapy. In that series, mean LogMAR VA of 0.26 was achieved at 24 months [3]. Furthermore, while anti-VEGF therapy has been shown to be effective [1], it is also associated with both ocular and systemic risks [4] such as endophthalmitis and cerebrovascular accidents. For patients who are unwilling to accept the risk of anti-VEGF therapy, we believe that PDT is a viable alternative to consider [3]. The authors also mentioned that switching from TimeDomain OCT (TD-OCT) (Stratus OCT, Carl Zeiss Meditec Inc., Dublin, CA, USA) to Spectral-Domain OCT (SD-OCT) (Cirrus OCT, Carl Zeiss Meditec Inc.) was the main limitation in excluding Central Subfield Thickness (CST) as a secondary outcome measure [1]. Several studies have described manual adjustment of the SD-OCTsegmentation lines [5] or the use of conversion equations [6] that reliably achieved retinal thickness measurements comparable to that of TD-OCT. While those studies compared Stratus OCT with Spectralis OCT (Heidelberg Engineering, Heidelberg, Germany), adjustment of segmentation boundaries are possible on the Cirrus OCT. It would be interesting see how the adjusted CST measurements correlate with the reported visual outcomes, and it may be possible to identify other prognostic factors identified for retinal and choroidal imaging [7–9], especially with the introduction of swept-source OCT [10]. In summary, we congratulate the authors on their interesting results, which add to our knowledge of the understanding and optimal management of myopic choroidal neovascularisation.

Optimal area of retinal photocoagulation necessary for suppressing active iris neovascularization associated with diabetic retinopathy

Dr. Tan receives research support from the National Healthcare Group Clinician Scientist Career Scheme Grant CSCS/12005. Dr. Tan receives travel support from Bayer, Novartis, and Heidelberg Engineering. Dr. Hariprasad is a consultant or on the speakers bureau for Alcon, Allergan, Bayer, Clearside Biomedical, Optos, Ocular Therapeutix, OD-OS, and Regeneron. In a recent article, Shiraya et al. [1] reported that the optimal percentage area of retinal photocoagulation required to suppress active iris neovascularization secondary to proliferative diabetic retinopathy was at least 50 % of the visible retina. While the results reported by Shiraya et al. are promising, we would like to highlight that there are large regions of retina beyond the area that was assessed by the authors, which have been shown to contribute to retinal diseases. The authors based their assessment of retinal area on a montage of overlapping fundus images. This appears to be similar to the 7-field montage used by the Early Treatment Diabetic Retinopathy Study (ETDRS), which images a region comprising approximately 75 degrees of the posterior pole. Recent advances in ultra-widefield retinal imaging allows a 200-degree field of view of the posterior pole in a single image [2], which is much larger than those obtained using conventional imaging. Studies have reported that peripheral pathology missed on 7-field imaging may be detected on widefield images [3]. Using ultra-widefield fluorescein angiography, ophthalmologists are able to visualize regions of retinal non-perfusion in the periphery of the retina and several studies have reported significant amounts of peripheral retinal non-perfusion in various retinal vascular diseases [4, 5]. The extent of retinal nonperfusion had some bearing on the disease severity and response to treatment. Neovascularization is mediated by vascular endothelial growth factor (VEGF), which is produced in response to ischemia [5]. It has been suggested that there is an association between peripheral retinal nonperfusion and the occurrence of retinal neovascularization as well as macular edema. As a result, it has been suggested that treating areas of ischemia with targeted retinal photocoagulation could reduce the VEGF load, while sparing the areas of perfused retina [2, 5]. M. C. Chew C. S. Tan (&) National Healthcare Group Eye Institute, Tan Tock Seng Hospital, 11 Jalan Tan Tock Seng, Singapore 308433, Singapore e-mail: Colintan_eye@yahoo.com.sg