Portal de Periódicos da CAPES

Screening for cystoid macular oedema in children with uveitis using the retinal thickness analyser

2008; Wiley; Volume: 86; Issue: 3 Linguagem: Inglês

10.1111/j.1755-3768.2007.01009.x

1755-3768

S. Maca, Gerhard Garhöfer, Christopher Kiss, Talin Barisani‐Asenbauer,

Retinal and Optic Conditions

Purpose: This study aimed to investigate the retinal thickness analyser (RTA) as a means to identify the presence of cystoid macular oedema (CMO) in children with uveitis, whether the course of CMO can be monitored using this method, and whether there is a trend towards a correlation between macular oedema and visual acuity (VA) in children. Methods: This prospective, cross-sectional study with observer-blinded analysis included 25 eyes. Standardized testing for best corrected distance VA (d-VA), near (reading) VA (n-VA) and slit-lamp examination were conducted. Using the RTA, a 3 × 3-mm scan of the macula was obtained, which was then used to discern CMO and calculate mean foveal thickness (MFT). Results: Macular scanning was possible in all children. Cystoid macular oedema was discerned in 10 eyes (40%) and ruled out in 15. In CMO eyes, d-VA was 0.5 Snellen and n-VA was 2 Jaeger; neither result differed significantly from those in eyes without CMO. Mean foveal thickness correlated with n-VA (r = 0.511, p = 0.015), but not with d-VA (r = 0.271, p = 0.191). After 3 months of tailored therapy, CMO was still detectable in six eyes. Changes in d-VA in the CMO and non-CMO groups were 3 ± 2.1 and 0.8 ± 1.8 Snellen, respectively; changes in n-VA were 1 ± 1.4 and 0.1 ± 0.3 Jaeger, respectively. Changes in MFT were − 244.8 ± 137.4 µm and − 0.8 ± 18.1 µm, respectively. A statistically significant correlation was found between the changes in MFT and n-VA (r = 0.629), but not with that in d-VA (r = 0.292). Conclusions: We used the RTA to establish the presence or absence of CMO according to measurements of the macular region. Our findings show that CMO is a common complication in children with uveitis and can be present even in cases with good d-VA. Mean foveal thickness as measured with the RTA correlates indirectly with n-VA. Cystoid macular oedema (CMO) is a serious complication of juvenile uveitis and a significant cause of visual impairment. Although CMO-complicated uveitis in children appears to attract less attention than it does in adults, early and adequate treatment of this potentially sight-threatening complication is crucial (Cunningham 2000; Smith 2002). However, early diagnosis of CMO among children is clinically challenging for several reasons. Firstly, the affected child may appear to be asymptomatic despite significant ocular inflammation. Secondly, children are often unable or unwilling to offer useful information about present symptoms, and, thirdly, clinical examinations and fluorescein angiography present technical challenges in this patient population. Therefore data, about the incidence and prevalence of CMO in children with uveitis are quite difficult to obtain and no gold standard for the clarification of CMO in children has been defined. The earliest reports of complications in childhood uveitis, published in the 1960s, do not even list CMO as an important complication (Cross 1965; Kazdan et al. 1967). The latest reports describe the development of CMO over time as affecting 17–49% of children with uveitis, with rising cumulative rates of complications in line with duration of disease (De Boer et al. 2003; Rosenberg et al. 2004; Mingels et al. 2005). The retinal thickness analyser (RTA) allows the clinician to visualize cross-sections of the retina in a rapid, simple and non-invasive manner. It is a scanning laser imaging device that yields thickness and contour information about the retina and optic nerve head (Zeimer et al. 1996). The instrument is currently used in the evaluation of macular oedema and assessment of the optic nerve head, both mainly in adult populations. There are relevant differences in the method of operation compared with that in optical coherence tomography II (OCT II): within one measurement, the RTA generates 16 adjacent cross-sections of the retina (Fig. 1), whereas the OCT II produces a single scan and a larger macular area can be covered only by performing multiple adjacent scans. As an additional advantage, the RTA includes a live fundus video camera, which allows for the survey of correct target fixation. During scanning the RTA then acquires and also saves a photographic fundus image on which the area scanned is indicated. The exact position of each individual scan can be located in this picture. This enables the examiner to identify the fovea even when patient compliance is low and/or correct fixation is not maintained. Scans of the macular area as displayed by the retinal thickness analyser. (A) Scan of a patient with good fundus visibility. (B) Scan of a patient with vitreous haze caused by vitritis and clearly visible intraretinal cystoid spaces caused by cystoid macular oedema. This work was designed to study the RTA as a means of identifying the presence of CMO without exposing patients to the difficulties and risks inherent in fluorescein angiography or angioscopy (Morgan & Franklin 1984). A secondary purpose was to find out whether monitoring the course of CMO in children is possible with this instrument. A rapid, non-invasive method of identifying and following CMO would be of great benefit to clinicians. The RTA (Talia Technology Ltd, Neve-Ilan, Israel) comprises a laser slit biomicroscope and digital camera attached to an ophthalmic table, a patient headrest, and a personal computer (software version 4.075). For a single measurement, a green helium-neon laser light with a wavelength of 543 nm is moved across the retina to produce 16 discrete slit images within a 3 × 3-mm retinal area. The reflected slit images are recorded digitally. Retinal thickness is derived from the separation between the anterior (corresponding to the internal limiting membrane) and posterior (corresponding to the retinal pigment epithelium) reflectance interfaces for 16 points along each slit via densitometry. The central scan area covers the macular region (Fig. 1). As an option, four overlapping measurements (each counting 16 scans) can then be taken to produce a composite RTA map covering the majority of the posterior pole of the retina. Depth resolution and depth precision are reported to be 5–10 µm and 50 µm, respectively. An internal fixation target simplifies patient fixation. A detailed explanation of the RTA optical principle has been described elsewhere (Zeimer et al. 1996; Gilmore & Hudson 2004). This prospective, cross-sectional study with observer-blinded analysis was carried out between September 2002 and February 2003 at the Department of Ophthalmology and Optometry, Medical University Vienna. It was approved by the university's institutional review board and conducted in accordance with the Declaration of Helsinki. Informed consent was obtained from each accompanying parent. Children who were referred to our department for the first time were included if they were aged under 14 years and diagnosed with either anterior or intermediate uveitis. Children with posterior uveitis were excluded. Data obtained at the baseline examination (visit 1) and a follow-up examination after 3 months (visit 2) were recorded and analysed. The sample comprised 25 eyes of 17 children aged 3−13 years. The anatomical location of uveitis was classified according to International Uveitis Study Group criteria (Bloch-Michel & Nussenblatt 1987). The children underwent standardized testing for best corrected distance visual acuity (d-VA) using Snellen charts at 6 m and near (reading) visual acuity (n-VA) using Jaeger reading charts at 30 cm in children with reading capability, and slit-lamp examination (including fundus examination through dilated pupils with a 90-D lens). Retinal scanning with the RTA was performed through dilated pupils, interpreted by one examiner (TB-A) and evaluated a second time by a blinded observer (SMM). Cystoid macular oedema was defined as the presence of clearly defined, hyporeflective, intraretinal cystoid spaces (Markomichelakis et al. 2004; Tranos et al. 2004) detected by both observers. When CMO was diagnosed, adequate therapy tailored to the patient's disease was prescribed. Statistical analysis was carried out using spss Version 12.0.1 (SPSS Inc., Chicago, IL, USA). Summary statistics of not normally distributed variables are reported as medians and ranges. Non-parametric tests were used to calculate differences between groups (Mann–Whitney U-test) and Spearman's correlation coefficient was used to analyse the correlation. The cut-off level for statistical significance was set at p < 0.05, two-tailed. In total, 25 eyes of 17 children aged 3−13 years (mean ± standard deviation [SD] 8 ± 3 years) were examined. Seven children suffered from anterior uveitis and 10 from intermediate uveitis (Table 1). Fundus examination with the slit-lamp was possible in all but two children (two eyes). Macular scanning was possible in all children. Cystoid macular oedema was identified by RTA in 10 eyes (40%). Normal foveal anatomy was confirmed in 15 eyes (60%) and CMO could thus be ruled out. Figure 1 demonstrates a typical reading in a subject with decreased fundus visibility and CMO. Distance VA was obtained in 22 and n-VA in 21 of 25 eyes. Data for eyes with CMO are shown in detail in Table 2 and 2, 3. Mean d-VA was 0.5 Snellen lines (range 0.1–0.8) in the group of eyes with CMO and 0.7 Snellen lines (range 0.2–1.0) in the group without CMO (p = 0.1). Mean n-VA was 2 Jaeger lines (range 1–4) in the group of eyes with CMO and 1 Jaeger line (range 1–12) in the group without CMO (p = 0.1). Distance visual acuity (d-VA) at visits 1 and 2 in eyes with cystoid macular oedema. Near visual acuity (n-VA) at visits 1 and 2 in eyes with cystoid macular oedema. Mean foveal thickness was calculated for each eye. As expected, values obtained were higher for eyes with CMO than without. Data are shown in Table 2 and Fig. 4. Mean foveal thickness indirectly correlated with n-VA (r = − 0.511, p = 0.015, Spearman's correlation), but did not correlate with d-VA (r = − 0.271, p = 0.191). Mean foveal thickness (MFT), in µm, at visits 1 and 2 in eyes with and without cystoid macular oedema (CMO). After 3 months of tailored therapy for all patients diagnosed with CMO, the condition was still detectable in six eyes. No additional cases were observed among those eyes without CMO at initial presentation. The change in d-VA showed 3 ± 2.1 Snellen lines of improvement in the CMO group and 0.8 ± 1.8 lines in the non-CMO group (p = 0.01). The change in n-VA was 1 ± 1.4 Jaeger lines in the CMO group and 0.1 ± 0.3 Jaeger lines in the non-CMO group (p = 0.02). Data for eyes with CMO are shown in detail in Table 2. The change in MFT was significantly higher in the CMO group (− 244.8 ± 137.4 µm) than in the non-CMO group (− 0.8 ± 18.1 µm) (p ≤ 0.0001, Mann–Whitney U-test). We found a statistically significant correlation between change in MFT and change in n-VA (r = 0.629, p = 0.004), but not with d-VA (r = 0.292, p = 0.157). In adults with intraocular inflammatory disease, CMO is known as a common cause of visual impairment (Rothova et al. 1996), with subsequent social and economic impact. The consequences of macular oedema in the paediatric population might have even greater effect on social and economic development. It is therefore important to identify the frequency of presentation of CMO in this population and to establish methods for the early and accurate diagnosis of this significant complication. Over the past decade, several different technologies have been developed for the non-invasive measurement of retinal thickness (Konno et al. 2001; Thomas & Duguid 2004; Iester & Mermoud 2005) and assessment of retinal oedema (Degenring et al. 2004; Guan et al. 2004). In the present study we used the RTA, a non-invasive, scanning laser, retinal imaging device, to assess the retina of children suffering from uveitis for evidence of concomitant CMO. The RTA was originally designed to yield reproducible measures of retinal thickness in ocular diseases such as diabetes mellitus or retinal vein occlusion, conditions that primarily affect adult populations. The fact that the instrument is able to generate 16 scan images in under 200 ms allows assessments to be made in patients such as children, who present with limited compliance. Other commercially available technologies are able to identify retinal oedema and cystic lesions but do not acquire regional scan data as rapidly as the RTA (e.g. OCT II) or represent cost-intensive techniques (e.g. OCT III). The short examination time and the fact that one scan can record 16 adjacent slit images helped enormously in maintaining compliance levels in the children in this study. Correct scan placement was facilitated by monitoring fixation throughout the scan. In this study population, 40% (10/25) of eyes presenting with intermediate or anterior uveitis were identified as having CMO via RTA analysis, although VA and clinical examinations performed by experienced observers did not suggest a diagnosis of maculopathy. These data are consistent with those reported by Rosenberg et al. (2004) (35%). De Boer et al. (2003) and Mingels et al. (2005) reported CMO rates of only 17% (3 years follow-up) and 16% (9–29% according to anatomical group, 5 years follow-up), respectively, pointing out the wide distribution of observations. Fundus examination using a slit-lamp and near vision testing was not possible in two children because of limited compliance; however, RTA imaging was possible in all children. After 3 months, therapy brought about resolution of CMO in four eyes and reduction in four. In two eyes CMO was refractory to treatment. Visual function as assessed by Snellen VA was maintained in this population, despite identification of CMO by RTA. There was no statistically significant association between the presence of CMO and distance Snellen acuity. Neither Snellen distance nor Jaeger near VA measures suggested the presence of CMO in the 40% of patients with RTA-indicated CMO. We propose that CMO as a complication of anterior and intermediate uveitis may be underdiagnosed in this population, given that neither Snellen acuity testing nor clinical examination were able to identify or rule out cases of CMO in concordance with RTA. With respect to the relationship between MFT and visual function, our data demonstrate a significant relationship with n-VA but not with d-VA. We are aware of the small sample size in this study and stress that our study was exploratory in nature. However, our findings support the data recently published by Kiss et al. (2006), which suggest that an n-VA task such as reading is more affected by CMO than is d-VA. Further limitations deserve mention. In several previously published reports the RTA was shown to be sensitive to opaque media, yielding overestimates of retinal thickness, as in eyes with severe vitritis (Neubauer et al. 2001; Polito et al. 2002; Iester & Mermoud 2005). In our study, as in other reports, vitritis and its associated media opacity resulted in reduced scan quality, as did iridolenticular synechiae. However, the main outcome measure of this study was the identification of retinal cystic spaces, not retinal thickness. The existence of cysts was more than clearly visible, when present (Fig. 1). In addition, standard values for retinal thickness in children as measurable by the RTA are not available to date, thus making interpretation dependent on the examiner's experience. We believe that it is reasonable to consider the RTA as a cost-saving screening tool for the presence of CMO and for evaluating CMO response to therapy. As neither d-VA nor clinical examinations were reliable indicators of CMO according to the RTA scans, and as it is difficult to interpret acuity in the presence of uveitis, we suggest that RTA scanning be used as an adjunct to standard clinical examination in children with uveitis. Parts of this work were presented at the Second Meeting of the International Society of Imaging in the Eye, Fort Lauderdale, FL, USA, in April 2004.