T his virtual special issue of the AJO ( https://www.sciencedirect.com/special-issue/10ZLM7LSRZ6 ) brings together types of scientific publications that might otherwise be overlooked in a high-impact journal focused on original research. This is partially a proceedings from a conference held in Singapore, November 21 to 22, 2024. Original research by the 70 members of the Multimodal Imaging in Uveitis (MUV) task force of the International Uveitis Study Group (IUSG) presented consensus-based diagnostic criteria for noninfectious posterior uveitis (NIPU) based on multimodal imaging. ,,,, Real-time polling of conference participants validated minimum imaging sets needed to extend diagnostic certainty beyond color fundus photography. Many of the other contributions would be classified as perspectives or editorials. ,, During the conference these expert opinions supplied the background to support the diagnostic criteria and now appear in print to educate others . A systematic review of the literature tabulated all published diagnostic features of NIPU and assessed their clinical utility.
Few types of retinal inflammations are as apt for an exploration of retinal imaging characteristics as the outer retinal inflammations. The so-called white dot syndromes have been kicking around as favorite topics of conferences and papers since these disparate diseases coalesced around the original multifocal evanescent white dot syndrome (MEWDS), which actually does resemble white dots or spots. The other disorders—acute posterior placoid multifocal pigment epitheliopathy (APMPPE), birdshot chorioretinitis (BSCR), multifocal choroiditis/punctate inner choroidopathy (MFC/PIC), and serpiginous choroiditis (SC)—share a feature of demarcated contrast with the fundus background with MEWDS but aren’t white or particularly dot-like. Multimodal imaging amplifies the differences between the disorders and supports revising their grouping as “white dots.” Masquerades include tuberculous uveitis, Toxoplasma chorioretinitis, syphilis, primary vitreoretinal lymphoma and a variety of other infections and macular degenerations as well as other noninfectious posterior uveitides such as acute zonal occult outer retinopathy (AZOOR). Somewhat ironically, even as an attempt is made to break down the artificial grouping as white dot syndromes, there is experimental evidence of immunoreactivity of the retinal pigment epithelium that may link the outer retinal inflammations to each other.
Most uveitis specialists will not use the term white dot syndromes except to convey to other ophthalmologists what they are talking about. The authoritative classification criteria, published in this Journal in 2021, used the names of each member of the classic quintet/sextet, but not the overall designation of “white dots.” The best evidence that “white dot syndrome” is an obsolete concept is based on objective retinal imaging as explored in the Singapore conference. Six imaging modalities were explored: the appearance on exam or color photography (looks still matter); characteristics of autofluorescence of the diseased or perilesional areas; staining, leakage, or blockage on wide-angle fluorescein angiography; hyper- or hypofluorescence on indocyanine green angiography; changes in thickness or structure on OCT; and abnormal vascularity or en face structure on OCTA. The latter two modalities in particular are increasingly defining the characteristics of posterior uveitis. , The capabilities of color photography and autofluorescence are also expanding with ultrawide imaging and selective color channeling with combinations of red, green or blue wavelength depending on application. Superior technology is now widely available to many ophthalmologists to objectively define the disorders and handily discriminate between them.
Both the conference and the papers in this VSI are poised to update the 2021 Standards of Uveitis Nomenclature II classification guidelines (SUN II). In SUN II, 101 investigators submitted large numbers of detailed case histories that were used to develop criteria for 25 named uveitic entities that became diagnostic orthodoxy. Images were optional and nonstandardized. A nominal group technique to validate cases, followed by machine learning to detect discriminatory features, was based on case summaries, but not the images themselves. A formal Delphi process approved the final criteria. The imaging characteristics in the final criteria for NIPU are limited. Only fluorescein angiography in acute phases was accepted as part of the definition of APMPPE. ICG angiography was needed in BSCR only if spots were not visible. No imaging, only size, location and appearance of lesions, was required for multifocal choroiditis with panuveitis and punctate inner choroiditis. , MEWDS was defined by the evanescent wreath-like lesion on fluorescein angiography or by outer retinal hyperreflectivity on OCT. Fluorescein angiography and autofluorescence criteria for SC accurately described only certain phases of the disease. The scantiness of defined imaging characteristics reflected methods that used fundus appearance to support diagnostic opinions but did not analyze the images themselves. By the end of 2017, when data collection was essentially complete for SUN II, there were one or more technical imaging findings, often repeatedly replicated, for each of the 6 modalities for each NIPU defined in SUN II. Publication of imaging findings expanded from 2018 onward. The evidence and consensus-based imaging guidelines presented in this issue now link the SUN II criteria to modern imaging knowledge.
Critics of this process will point to the small number of cases (fifteen) that the committees used to define the imaging characteristics of NIPU. ,,,, Nominal group techniques assured that only classic cases were considered and that the imaging was of the highest quality. An extension of this critique would be to point out that deep machine learning with hundreds or thousands of cases might be more reliable in discriminating imaging features than a group of humans. Humans are still in charge of the truth and in this case the judgment, starting with the SUN II criteria, was that these were actual examples of real entities. The final product works organically with the way ophthalmologist evaluate their patients and restrains the wasteful impulse to order all 6 modalities on each patient by recommending minimum imaging sets. We should still seek the fine detail the human eye missed that might hold the clue to the pathology of these disorders and research new and better imaging techniques to help us “see” our patients. McKay et al. outline the challenges of using AI to interpret imaging in ocular inflammatory diseases and the steps forward if such a project is undertaken.
The 70 members of the MUV task force worked for more than 2 years to complete this project without grant funding. Future work will include supplements to the other entities defined in SUN II that would benefit from updating with a fresh look at what retinal imaging offers in the classification and diagnosis of uveitic entities.
CRediT authorship contribution statement
Janet L. Davis: Writing– review & editing, Writing– original draft, Conceptualization.
Funding/Support: This study received no funding.
Financial Disclosures: Data Safety and Monitoring Boards for 4D Molecular Therapeutics, Kodiak Biosciences, Aura Biosciences. All authors attest that they meet the current ICMJE criteria for authorship.
Other Acknowledgments: The dedication of the members of the Multimodal Imaging in Uveitis (MUV) taskforce who worked tirelessly to complete this project.
This invited editorial introduces the Virtual Special Issue on Multimodal Imaging of Non-Infectious Posterior Uveitis by The International Uveitis Study Group MUV Working Group.
References
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
