The diagnosis of CMVR is not tricky basing on the clinical appearance of the fundus and documentation of an immunodeficient status. Even nonophthalmologists can achieve it through retinal images or binocular indirect ophthalmoscopy for patients with AIDS. However, as a treatable disease, delay in the diagnosis and treatment of CMVR may lead to irreversible loss of vision and systemic deterioration. In our study, 8 (8.7%) out of 92 patients were misdiagnosed. Six patients consulted the ophthalmology department first for the ocular symptom before the diagnosis of AIDS, and one patient concealed the history of HIV infection. When their diagnosis of AIDS was clear, the blood CD4 + T lymphocyte was extremity low, which means the body’s immune state is very poor and progressed to an advanced stage. Seven patients in this study diagnosed with AIDS showed an extremely low level of CD4 + T lymphocyte, ranging from 1 cells/ul to 9 cells/ul. The delay of HIV infection may lead to serious consequences: compromising timely management and enhancing propagation of the epidemic in the country.
Besides, the CMVR lesions in this study had deteriorated when final diagnosis was clear. All eyes presented binocular involvement. One patient is blind, and two patients had a low vision when the diagnosis is precise. Four eyes presented pan-retinal involvement. Fourteen eyes had optic disc or macular area involved, which is associated with poor visual prognosis.
Some misdiagnoses were partly attributed to the underrated systemic symptoms insulting at the ophthalmology department because four patients developed variable extraocular manifestations before or in the course of the disease: abdominal pain, diarrhea, oral leukoplakia, rash, itchy skin, nausea, and vomiting, which prompted systemic diseases. However, since the visual dysfunctions are the main problem and, therefore, qualified patients for permanent disability, they had concealed these extraocular symptoms from those involved in their care.
CMV infection reaches the eye through hematogenous spread. Histologically, CMVR appears as areas of full-thickness retinal necrosis and edema or exudative detachment. Clinically, it has a young male predominance, as showing in this study. There are various findings of the fundus in CMVR, not just “cheese and ketchup” changes. It has various clinical classification according to the clinical fundus forms, Standardization of Uveitis Nomenclature working group descriptors, or the involvement of immune reconstitution inflammatory syndrome (IRIS). There are four recognized clinical forms, including typical form, edematous; atypical form, indolent; perivascular form; and optic neuropathy. Among six patients (12 eyes) with CMVR in this study: 5 eyes presented typical form with yellowish-white retinal lesions and retinal hemorrhages; 4 eyes presented optic neuropathy; 4 eyes presented opaque white granular retinal lesions with no hemorrhages, which could be easily misdiagnosed by inexperienced doctors; 1 eye presented both optic neuropathy and perivascular form.
These yellowish-white lesion and retinal hemorrhages were characteristic manifestations but could be easily confused with diabetic retinopathy, retinal vein occlusion, renal retinopathy, intraocular lymphoma, Coats' disease or ocular toxoplasmosis. The seventh patient showed quadrantally distributed white granular lesions without retinal hemorrhages and was misdiagnosed as PORN, which is also a kind of necrotizing herpetic retinopathies detected in immune-compromised patients. CMVR and PORN might present confusing fundus changes, which could be difficult for inexperienced doctors, but they are different in the aetiological agent, treatment, and prognosis.
The eighth patient was initially misdiagnosed as CMVR, because of the yellowish-white lesion and retinal hemorrhages around the optic disc. With HIV-infection finally excluded, chronic renal insufficiency confirmed, and retinal hard exudates near the macular area, the patient was finally diagnosed with renal retinopathy. The critical distinguishing signs of CMVR and renal retinopathy is retinal hard exudates near the macular area, which is rarely seen in CMVR. Other confusing features were the signs of inflammatory reaction in the anterior segment or vitreous body, which may be misleading to the diagnosis of iritis, or Behcet's disease. For ambiguous cases, PCR analysis from aqueous or vitreous specimens can be helpful to different infected and non-infected diseases, and find the pathogen.
HIV infection might be firstly diagnosed in the ophthalmology department because of CMVR. An ophthalmologist should pay more attention to the diagnosis of CMVR to avoid delayed ocular or systemic treatment. In our study, the rate of misdiagnosis was 8.7%. However, this retrospective study suffers from certain limitations without long-term effective follow-up.