Article Text

Download PDFPDF

Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.


From questions on page 1599


Posterior segment infections in HIV-infected patients present as a necrotising retinitis or choroiditis, categorised by the CD4+ count. Acute retinal necrosis (ARN), syphilis, toxoplasma and cryptococcus occur with higher CD4+ counts, whereas CMV and progressive outer retinal necrosis develop at lower CD4+ counts.1 In this case, a diagnosis of presumed serological negative CMV retinitis was deemed unlikely in the background of high CD4+ counts. However, this diagnosis was not completely excluded. Immune restoration inflammatory syndrome was unlikely, as the patient had not previously undergone HAART by the time of the initial presentation.


In January 2006, consultation with a national ocular HIV centre was undertaken, and it was suggested that the chronic course of retinitis together with signs of vascular attenuation may both indicate CMV retinitis. Brain imaging and lumbar puncture were not performed at this stage, due to the absence of neurological symptoms and signs. A trial of anti-CMV therapy was subsequently commenced in consultation with the patient. A vitreous biopsy was done, and a 2-week course of intravitreal injections of ganciclovir 2 mg/0.1 ml was commenced. Oral valganciclovir was initiated to reduce the risk of retinitis in the fellow eye and prevent extraocular involvement. The clinical appearance did not change after anti-CMV treatment (fig 1A), and the vitreous biopsy was PCR-negative for CMV, VZV and HSV.

In May 2006, the CD4 count was 210 cells/μl, and the hospital HIV service started HAART. There was a good response to HAART with a viral load reduction from 200 000 to 670 copies. In July 2006, the retinitis had resolved (fig 1B), with a mild vitritis persisting. A 1-month course of oral prednisolone was started, and the vitritis resolved almost completely.

Nine months after the initial presentation, the patient reported a temporal subjective scotoma and generalised headaches. VF testing demonstrated a complete left-sided homonymous hemianopia. An MRI scan was later confirmatory of the underlying diagnosis (fig 2). A lumbar puncture was subsequently performed, and JCV DNA was detected on PCR testing.

Further ocular treatment involved a maintenance course of topical prednisolone twice-daily over the following 6 months. There were no further recurrences of retinitis. At a recent visit in August 2007, the patient was asymptomatic with no signs of ocular inflammation (fig 1C) and normal vision. The CD4+ count remained above 220 throughout.


There are extensive high signal changes seen within the right occipital lobe extending into the right parietal and temporal lobes, with mild cortical atrophy (fig 2). The features are consistent with a diagnosis of progressive multifocal leucoencephalopathy (PML). The diagnostic signal changes are visible during the flare sequence of MRI. The lack of enhancement excludes conditions such as toxoplasmosis and lymphoma. Despite the high signal changes, the lack of mass effect would make an infiltrative astrocytoma unlikely. CMV encephalitis is part of the differential, but there is no sign of subependymal enhancement to suggest CMV ventriculitis.


Our case explores diagnostic dilemnas and highlights the difficulties in reaching a diagnosis in a patient with retinitis. At the initial stage, causes of impaired cellular immunity including the possibility of HIV infection need to be explored. Both serological and ocular sampling to investigate an infective aetiology are important in HIV-positive patients. In the presence of negative investigations, it is important to collaborate with ocular HIV and HIV medical specialists in order to guide subsequent management. Non-invasive VF testing allows management to be directed towards radiological investigation, and CSF sampling may provide confirmatory results. JCV is a human polyomavirus2 and is well recognised to cause the central demyelinating condition PML.3 Ocular inflammatory disease is previously unreported with JCV. Our case demonstrates that presumed JCV retinitis occurs with a high CD4+ count and may precede the development of PML.

In HIV-infected patients, CD4+ counts are important for investigation and treatment strategies. CMV retinitis develops at CD4+ counts less than 50 cells/μl, and our patient underwent complete serological, aqueous and vitreous testing to exclude an infectious aetiology. The diagnosis of presumptive JC viral retinitis was made, as the ocular fluids were unable to be tested for JCV. A retinal biopsy was not performed. Furthermore, there was no response to a trial of anti-CMV therapy. In our patient, this diagnosis had been thoroughly excluded.

JCV is ubiquitous by the presence of antibodies to the virus in approximately 90% of the normal adult population.4 JCV has been shown to be transactivated in vitro by either CMV or HIV-1.5 HIV infection of ocular tissue is a prerequisite to the viral transactivation. The level of JCV DNA may be significantly increased in HIV-infected patients with a history of CMV retinitis, but our patient did not demonstrate any CMV positivity. The site of JCV ocular infection may be the retinal astrocytes, and lymphocytes may act as carriers of JCV to the eye. Currently, JCV PCR testing of ocular tissue is not a routine test used in clinical practice. Non-infectious HIV retinopathy (HIVR) is associated with focal retinitis, cotton wool spots and haemorrhages.6 However, the sensitivity of JCV detection in ocular tissue may be unpredictable in these cases, since HIVR is a mild vasculopathy with small amounts of virus present within the retina.7 PCR testing of cerebrospinal fluid (CSF) is a sensitive diagnostic test for JCV infection and a reliable virological marker of PML disease activity.8

HIV-1 has been shown experimentally to upregulate JCV protein expression in glial cells,5 and both HIV-1 and JCV may infect the same cells, that is astrocytes in the eye and brain.10 Primary JCV-induced PML may be as a result of reactivation of latent infection in immunocompromised individuals by HIV-1. Therefore, it is possible that the effects of HAART on the JCV lytic cycle may be locally mediated by HIV-1. The HIV load is reduced following HAART, and this has been associated with a reduction in the JCV CSF load.8 It is thought that the PML may stabilise as the JCV load reduces post-HAART. With regard to ocular disease, the reduction in JCV viraemia following HAART may lead to inhibition of retinal astrocyte destruction, and remission of retinitis and uveitis. During the 9-month active phase of the retinitis in our case, the JCV-induced PML may have been initially subclinical. The persistent retinitis and JCV-induced PML progression is consistent with the delayed onset of headaches and visual-field defect, followed by the later radiological signs of PML. After HAART, the radiological signs of PML did not progress over 6 months, and the headaches resolved completely.

An important finding in our patient was the temporal development of PML. On review of VF tests, clinical PML developed 9 months after the initial onset of retinitis. The retinitis and HIV load responded favourably to the HAART, and this treatment response follows the increased survival of PML patients recognised after HAART.9 JCV DNA levels in CSF may be monitored after HAART to guide management.

The recommended approach to the management of HIV-infected patients with normal/or high CD4+ counts presenting with retinitis is to investigate serology, ocular investigations and CSF sampling. If all investigations are normal for CMV, VZV, HSV, toxoplasma, cryptococcus, syphilis and Lyme disease, then JCV infection should be considered. Ocular aqueous and vitreous samples, followed by CSF samples, should all be tested for JCV DNA by PCR. A second option is to perform a retinal biopsy, followed by electron microscopy and immunohistochemical analysis. Baseline VF testing and CSF sampling have a role in the initial investigations performed by ophthalmologists and HIV clinicians in order to detect signs of PML. In our case, the PML has not progressed 1 year after the radiological diagnosis was made. The patient remains under ocular and HIV surveillance.


Presumed Jamestown Canyon viral retinitis and associated progressive multifocal leucoencephalopathy


Linked Articles