Sir,

We thank Dr Agrawal1 for his correspondence, which helpfully highlights the important role of infectious pathogens in the aetiology, diagnosis, and treatment of uveitis. Although the title of our review explicitly limits its scope to non-infectious intraocular inflammation, we endorse his emphasis on the importance of excluding treatable infectious uveitides before embarking on treatment with immunosuppressive agents. However, while acknowledging wide global variation in the burden of infectious diseases, historical surveys endorse our assertion that in Western populations the balance has firmly swung in favour of autoimmunity and autoinflammation—in Guyton and Woods2 (USA), 1941 series, over 60% of uveitis cases were due to tuberculosis or syphilis, whereas half a century later, Rothova et al3 (Netherlands) reported that these infections accounted for less than 3% of their cohort.

It is also important to reiterate the potential crossover between infectious and autoimmune/autoinflammatory diseases, as outlined in our overview of immunopathology; in particular the role of molecular mimicry and the key contribution concomitant infection has in giving innate immune permission to the antigen-specific adaptive immune responses we observe in the eye. We also increasingly have access to molecular biological technologies, which enable us to exclude direct intraocular infection from aqueous or vitreous biopsies, and in the absence of a proven pathogen we strongly advocate appropriate immunosuppression to maintain vision.