Statistics from Altmetric.com
Editor,—Between 16% and 40% of patients with AIDS develop cytomegalovirus retinitis (CMVR).1 2 This typically presents when the CD4+ count drops below 75 cells/μl.3 Visual loss is usually due to retinal detachment, macular or optic nerve infection, or vascular occlusion. Typically a minimal inflammatory response is mounted. Patients with CMVR not associated with HIV infection may develop a vitritis with cystoid macular oedema (CMO) during immune restoration.4
With the advent of highly active antiretroviral therapy (HAART) patients with AIDS are enjoying improvement in their immune status, with falling HIV viral loads and rising CD4+ counts. Such patients may mount a significant inflammatory response against CMVR, causing a regression of the CMV infection but a paradoxical visual loss due to the inflammation.5 We describe two such patients with CMVR and vitritis, who lost vision as a result of CMO.
Patient 1 was diagnosed with zone 2 CMVR affecting the left eye in May 1996. He was induced on intravenous ganciclovir 5 mg/kg twice daily for 2 weeks and maintained on intravenous ganciclovir 5 mg/kg once daily for 5 months until progression of CMVR occurred. He was then induced on intravenous cidofovir 5 mg/kg/week followed by intravenous cidofovir 5 mg/kg twice a month maintenance. At the same time he was started on HAART therapy consisting of D4T, 3TC, and indinivir. There was a rapid resolution of active CMVR with a rise in CD4+ count and a fall in the HIV viral load (Fig 1). However, 2 months later his vision dropped to from 6/12 to 6/24 and vitritis, optic nerve swelling, and CMO were noted (Figs 2 and 3). Vision in the right eye was maintained with no sign of CMVR.
Patient 2 was diagnosed with zone 2 CMVR affecting the right eye in October 1996 (Fig 4). He was induced on intravenous ganciclovir 5 mg/kg twice daily for 2 weeks followed by maintenance on ganciclovir by mouth 3 g/day for 4 months until progression of the CMVR occurred into zone 1 in the right eye and zone 3 in the left eye. He was treated with bilateral ganciclovir implants with no systemic cover. Treatment of HIV with AZT and 3TC was continued. Immune recovery followed the successful treatment of pulmonary tuberculosis. In May 1997 his CD4+ count had risen to 250/μl with a low HIV viral load. Vision in the right eye fell from 6/9 to 6/24 due to macular oedema and mild vitritis. Treatment with topical prednisolone four times daily and with acetazolamide 250 mg twice daily, produced a sustained two line improvement in visual acuity (Figs 5 and 6). Vision in the left eye was maintained at 6/5 with stable zone 3 disease and no CMO.
The picture of CMV retinitis is changing with the advent of effective retroviral therapy. CMVR has recently been diagnosed in patients with relatively high CD4+ counts (197–270/μl) following the initiation of HAART.6 A syndrome of macular serous exudation7 has recently been described, and there are reports of iritis8 in patients on intravitreal cidofovir. However, only a few cases with CMO have been described.9 10
Both patients in this study developed macular oedema. In patient 2 use of a ganciclovir implant may have contributed to macular leakage. However, in both patients only the eye with extensive disease developed CMO.
Further analysis of the rate of immune recovery, viral load levels, and other immune markers may help to predict which eyes may become affected by CMO.
The question of treatment causes a dilemma. The systemic use of immunosuppressive agents may be contraindicated; however, the use of topical or sub-Tenon’s steroid may have some benefit. One of our patients benefited from acetazolamide 250 mg twice daily, although the side effects when combined with antiviral drugs made this treatment unsustainable. Others have used NSAIDS topically and systemically with some effect.
Cystoid macular oedema in HIV+ patients with CMV retinitis may represent a healing phase of the disease, which sadly still causes visual loss and presents physicians and immunologists with new clinical challenges.