Could a macular telangiectasia type 2-like phenotype characterize retinal JC virus infection?

We read with great interest the recent correspondence article by Hupin et al.[1] describing a case with ocular involvement of JC virus (JCV) infection. We very recently reported a case of JCV-related retinopathy for an individual with common variable immunodeficiency [2]. Clinically, our case presented with rapid progression of unilateral retinopathy involving all retinal layers [2]. JCV was detected exclusively in ocular samples – specifically in the aqueous humor, vitreous, and retina – by PCR and positive immunodetection of the SV40 antibody (which cross-reacts with JCPyV). Positivity was predominantly within the retinal ganglion cell layer, but was also detected in some nuclei of cells within the retinal inner nuclear layer. All other biological samples, including blood, urine, and cerebrospinal fluid, tested negative for the virus.

Regarding the diagnosis proposed in Hupin et al.[1], we agree that some multimodal imaging features are consistent with macular telangiectasia type 2 (MacTel 2). However, contrary to the authors’ suggestion, we believe that the angiographic finding of macular fluorescein leakage is not incompatible with MacTel 2. Indeed, macular fluorescein leakage has been described as a relatively common feature of MacTel 2 [3]. Concerning macular optical coherence tomography (OCT) findings, Hupin et al.[1] report asymptomatic outer retinal layer abnormalities 11 months after infection that do not evoke a MacTel 2 phenotype [3]. Briefly, features included unilateral widening and blurring of the interdigitation and ellipsoid zones and bilateral hyperreflectivity within the outer nuclear layer. Two-and-a-half years later, macular thickening was observed; again, not typical of MacTel 2 [4]. Subsequently, the authors describe bilateral retinal thinning with the presence of low reflective spaces. Interestingly and similarly to our reported case [2], these OCT abnormalities involved all retinal layers for the case reported by Hupin et al.[1]. On the other hand, the imaging features differed: a MacTel 2-like phenotype for Hupin et al.[1], versus diffuse retinal atrophy for our case [2]. A retinal biopsy was absent from Hupin et al.[1] and would have been helpful for confirming this rare manifestation of JCV infection in the retina. This means for now we can only speculate that ocular involvement of JCV could affect the retina and could manifest via different clinical presentations.

From a pathophysiological standpoint, MacTel 2 is thought to arise from early Müller cell dysfunction [5]. Müller cells reside in the retinal inner nuclear layer [6]. In our report [2], as mentioned above, we demonstrate JCV expression within this same retinal layer without actually identifying the specific immunostained cells. We therefore support the hypothesis put forward by Hupin et al.[1] suggesting a potential pathogenic role of JCV in the retina. Indeed, while it is possible that MacTel 2 developed coincidentally alongside ocular JCV infection in the case reported by Hupin et al.[1], it is also plausible that JCV infection actually have contributed to the development of the MacTel 2-like phenotype. This MacTel 2-like phenotype could thus specifically represent a distinct clinical entity directly attributable to JCV infection.

Overall, these two reports combined [1,2] reinforce the concept that the eye may serve as both a potential viral reservoir for JCV and a target organ, and that ocular JCV invasion may affect the retina differently. Indeed, could these two distinct clinical presentations result from infection with different JCV variants? Future reports on ocular involvement of JCV, particularly including long-term multimodal imaging findings, will be essential for deciphering the different retinal features that develop, as well as when they develop and following infection with which specific JCV variant. In the meantime, in clinical practice we aim to investigate JCV infection systematically when confronted with atypical forms of progressive retinopathy.

Acknowledgements

We thank Jacqueline Butterworth, PhD (Toulouse University Hospital) for help with manuscript drafting and language editing.

Conflicts of interest

There are no conflicts of interest.