Editor, Acute macular neuroretinopathy (AMN) is a rare disease of unknown aetiology (Bos & Deutman 1975; Turbeville et al. 2001). We have observed a case with acute onset during intravenous cytokine-releasing immunotherapy. A 25-year-old woman with primary oxalosis type 1 was liver and kidney transplanted at the age of 9 years and kidney transplanted again at the age of 24 years. She had no history of eye disease. The patient was treated with tacrolimus 2 mg BID, mycophenolate mofetil 1 g QD, deflazacort (a prednisone analogue) 9 mg QD, omeprazol 20 mg QD, calcitriol 1 μg QD and carvedilol 12.5 mg QD and an oral contraceptive when she was hospitalized 14 months after her most recent kidney transplantation with decreasing kidney function and acute humoral and cell-mediated graft rejection. She was immediately started on high-dose intravenous methylprednisolone and oral prednisolone. Twelve days later, a series of five daily plasmapheresis sessions was begun. The following day, an infusion of intravenous rabbit anti-thymocyte globulin 100 mg (1.5 mg/kg; Thymoglobulin®; Genezyme, Cambridge, UK) was administered at a rate of 25 mg/hr, preceded by intravenous methylprednisolone 250 mg, clemastine 2 mg and oral paracetamol 1 g. During the 4-hr-long infusion and 24 hrs thereafter, the patient suffered severe malaise but had no fever. The level of plasma c-reactive protein was increased from <1 to 55 mg/l (reference <10 mg/l). Within 12 hrs after the initiation of the anti-thymocyte globulin infusion, the patient developed blurred vision and dark spots centrally in the visual field of both eyes. Anti-thymocyte globulin therapy was interrupted the next day after the T-lymphocyte count had fallen below the target level of 50 per μl. Five infusions of anti-thymocyte globulin 75 mg given over the following 10 days were not associated with malaise or aggravation of the visual symptoms. On the 3rd day after beginning anti-thymocyte globulin infusion, we found best-corrected visual acuity (BCVA) 0.3 in the right eye and 0.8 in the left eye. Large pericentral relative scotomata were demonstrated in both eyes by automated perimetry. Sharply demarcated dark fundus lesions associated with photoreceptor layer hyporeflectivity on optical coherence tomography (Fig. 1) and areas of reduced amplitude on multifocal electroretinography corresponded to the scotomata in both eyes. A diagnosis of AMN was made (Bos & Deutman 1975; Turbeville et al. 2001; Monson et al. 2007). Given the benign nature of the ocular condition and the threat of graft rejection, no recommendation against continued anti-thymocyte globulin infusion was made. Subjective and objective remission of the visual field defects was seen from 1 week after the onset of symptoms. Infrared scanning-laser ophthalmoscopy (left) and transfoveal optical coherence tomography (right) showing dark macular lesions matched by hyperreflectivity of the outer nuclear layer and hyporeflectivity of the photoreceptor/retinal pigment epithelium complex. The abnormalities had largely disappeared 5 months later (lower). The patient was discharged in good shape 7 weeks after the initiation of anti-thymocyte globulin infusion, but with reduced kidney function, s-creatinine 260 μmol/l. Twelve months after the onset of AMN kidney function was stable, BCVA had improved to 1.25 in both eyes and only small paracentral relative scotomata were found. The fundus lesions and the areas of hyporeflectivity in the photoreceptor layer had shrunk but remained partially visible. The close temporal association between the anti-thymocyte globulin infusion and the onset of AMN suggests a causal relation. An initial infusion of anti-thymocyte globulin results in cytokine release that peaks after 4 hrs despite pre-dosing with prednisolone (Guttmann et al. 1997). This response includes marked increases in plasma interleukin 6 and tumour necrosis factor alpha (TNF-α) that are also key factors in producing the symptoms of influenza A and in mobilizing host defence (Hayden et al. 1998). Flu-like illness commonly precedes AMN (Turbeville et al. 2001). Interestingly, TNF-α has been shown to be involved in photoreceptor degeneration in retinal detachment (Nakazawa et al. 2011). Improved visual function and absence of malaise during repeated anti-thymocyte therapy are compatible with only the first infusion of anti-thymocyte globulin resulting in significant cytokine release (Guttmann et al. 1997). In conclusion, our observation suggests that cytokines released from T cells that disintegrated as a result of anti-thymocyte globulin infusion may have been be involved in the development of AMN.
Aims: To examine whether the addition of dorzolamide to timolol monotherapy influences oxygen saturation in the human retina. Methods: Non-invasive spectrophotometric retinal oximetry was used to measure oxygen saturation in retinal vessels. Twenty patients with open-angle glaucoma (11) and ocular hypertension (9) were recruited. The patients were randomised into receiving timolol monotherapy or dorzolamide–timolol combination for an 8-month test period, followed by a second test period, before which the patients switched treatments. Oximetry measurements were performed at 2-month intervals during each period. Of the 20 patients, 13 followed the study protocol into the second test period, and 10 managed all study visits. Results: The oxygen saturation in retinal vessels was stable within the test periods. The mean arteriolar saturation was 96 (2)% (mean (SD)) during timolol monotherapy and 97 (2)% during dorzolamide–timolol combination therapy (p = 0.17, all patients pooled, n = 13). Corresponding values in venules were 66 (5)% during timolol monotherapy and 65 (6)% during dorzolamide–timolol therapy (p = 0.13). Patients who started on dorzolamide–timolol combination showed a significant reduction in arteriolar (98 (2)% to 95 (2)%, p Conclusion: Adding dorzolamide to timolol monotherapy has a minimal effect, but going from dorzolamide–timolol combination to timolol alone lowered arteriolar and venular oxygen saturation. The retinal oxygen saturation measurements show a high degree of stability over an extended period of time. Previous studies have suggested increased retinal and optic nerve blood flow with dorzolamide. Unchanged oxygen saturation and increased blood flow would indicate increased oxygen delivery to the retina.
Abstract Purpose To assess whether the direct and consensual postillumination (ipRGC‐driven) pupil light responses to chromatic light stimuli are equal in healthy subjects. Methods Pupil responses in healthy volunteers were recorded using a prototype binocular chromatic pupillometer (IdeaMedical, Copenhagen), which is capable of both direct and consensual pupillometry measurements. The device uses a pair of dual monochromatic narrow bandwidth LED light sources, red (660 nm) and blue (470 nm). Pupil light responses were recorded with infrared video cameras and analysed using custom‐made circuitry and software. Subjects were randomized to receive light stimuli at either the right or left eye after 5 min of dark adaptation. Pupil light responses were recorded in both eyes for 10 seconds before illumination, during illumination and 50 seconds after illumination with red and blue light. Three variables were defined for the recorded pupil responses: the maximal constriction amplitude (CA max ), the pupil response during illumination and postillumination pupil response (PIPR). Results No difference was found in the pupil response to blue light. With red light, the pupil response during illumination was slightly larger during consensual illumination compared to direct illumination (0.54 and 0.52, respectively, p = 0.027, paired Wilcoxon's test, n = 12), while no differences were found for CA max or the PIPR. Conclusions No difference was found between direct and consensual pupil response to either red or blue light in the postillumination period. Direct and consensual responses can readily be compared when examining the postillumination pupil response to blue light as estimation of photosensitive retinal ganglion cell activation.
Purpose Disturbances in blood flow and oxygenation are believed to be involved in diseases such as diabetic retinopathy, retinal vascular occlusion and possibly glaucoma. The purpose of the studies presented here is to measure haemoglobin oxygen saturation (SatO2) in retinal vessels in patients with these diseases. Methods Our retinal oximeter is based on a fundus camera, which is coupled with beam splitters and narrow band-pass light filters. The oximeter yields fundus images with 4 wavelengths of light simultaneously. Two wavelengths, 605nm and 586nm, are used for estimation of SatO2. Measurements were made on (A) 7 patients with CRVO, (B) 19 patients before and after glaucoma surgery and (C) 21 patient with diabetic retinopathy, who were compared with 20 healthy volunteers. Results (A) Venous SatO2 was 53±9% in CRVO eyes and 65±4% in fellow eyes (mean±SD, p=0.015, n=7). (B) Glaucoma surgery has a minimal effect on SatO2 in both arterioles (2% rise, p=0.046, n=19) and venules (no change). IOP was lowered by 13mmHg. (C) SatO2 in retinal venules was 60±8% in healthy volunteers (n=20), 67±7% in patients with non-proliferative DR (n=12, p<0.05 compared to healthy) and 68±6% (n=9, p<0.05) in patients with proliferative DR after PRP treatment. Conclusion Human retinal vessel oximetry can detect changes in various ocular diseases and this may help us understand the pathophysiology. The decreased venous SatO2 in CRVO is probably caused by decreased blood flow. A small change in SatO2 after glaucoma surgery may indicate a large change in oxygen delivery since blood flow may be increased with IOP lowering. Higher venous SatO2 in DR patients may for example be explained with arteriovenous shunting of blood. Commercial interest
Abstract Purpose To investigate the effect of intravitreal anti‐VEGF on retinal oxygenation in patients with central retinal vein occlusion. Intravitreal anti‐VEGF injections have recently become an approved method of treatment for retinal vein occlusions, as several studies have shown a beneficial effect on visual acuity and retinal thickness. However, there are indications that intravitreal anti‐VEGF may cause constrictions of the retinal vessels, leading to a reduction in retinal blood flow. Methods Retinal oxygen saturation in patients with CRVO was analysed, using the Oxymap Retinal Oximeter P3, before and during 6 months of treatment with intravitreal injections of ranibizumab. Results Our preliminary results indicate that retinal oxygen saturation is not decreased by intravitreal anti‐VEGF injections. Commercial interest
To investigate the effect of intravitreal injections of the vascular endothelial growth factor inhibitor ranibizumab on retinal oxygenation in patients with central retinal vein occlusion (CRVO).
Methods
Retinal oxygen saturation in patients with CRVO was analysed using the Oxymap Retinal Oximeter P3, before and during 6 months of treatment with intravitreal injections of ranibizumab.
Results
At presentation, retinal venous oxygen saturation was lower in eyes with CRVO than in the healthy fellow eyes (32±13% vs 59±10%, respectively, p=0.001) whereas retinal arterial saturation was higher in eyes with CRVO than in the fellow eyes (95%±8% and 91%±3%, p=0.04). Mean visual acuity increased from 51±24 letters ETDRS at baseline to 66±24 and 69±20 letters ETRDS, respectively, at 3 months and 6 months treatment (mean±SD, p<0.0001, repeated measures analysis of variance) and central retinal thickness was reduced from 697±139 µm to 368±113 µm and 340±96 µm, respectively, from baseline to 3 months and 6 months treatment (p<0.0001). Venous saturation increased during treatment (from 35.5%±13.8% at baseline to 43.1%±10.8% and 43.5%±13.7% after 3 months and 6 months treatment, respectively, p=0.012), while no significant change was found in arterial saturation (p=0.24).
Conclusions
Retinal venous oxygen saturation was markedly reduced in untreated CRVO and was roughly halfway normalised during intravitreal ranibizumab treatment. Retinal artery oxygen saturation was not reduced in CRVO.
