Uveitis Classification System Offers a Step Toward Standardization

A classification system with criteria for the 25 main uveitis entities will not only support research efforts by facilitating homogeneous study cohorts — it can also create more standard disease definitions, allow for consistency between electronic medical record databases, and assist clinicians in diagnosing these complex ocular disorders which sometimes involve systemic reactions. 

In 2005, the Standardization of Uveitis Nomenclature (SUN) working group was established to develop criteria for the uveitides, and they completed a process of 4 phases: building informatics, gathering case records, selecting cases, and applying machine learning to training and validation sets. After best distinguishing characteristics were set for many of the inflammatory diseases, SUN was published in 2021. Researchers recently conducted a chart review that included 1,143 patients who sought care at the uveitis clinic of University of Colorado Hospital from 2013 to 2020, and the resulting data is published in American Journal of Ophthalmology. Participants were on average 47±20.1 years of age at first visit.

Investigators evaluated patient history, exam notes, imaging, and lab work to find cases which were classifiable — categorization did not have to be the same as clinical diagnosis. Of the total sample, 572 patients exhibited uveitides presenting clinical features not yet listed — thus researchers were not able to attempt classification according to SUN. For the 571 remaining, 522 could be classified with the criteria, and 94.3% of these also matched clinical diagnoses by physician specialists. Five individuals were evaluated for classification of 2 disorders; 1 of whom met criteria for both, 3 for only one of the conditions, and 1 did not meet either category.

Participant records for which classification could not be attempted consisted primarily of undifferentiated anterior uveitis classes, as well as iris inflammation; undifferentiated panuveitis; and retinal vasculitis. Other factors affecting results involved individuals who did not complete their diagnostic testing, and diseases that have yet to be described and entered, such as uveitis-glaucoma-hyphema syndrome.

Previous studies have suggested that listings in SUN will likely continue to expand along with improvements in imaging technology and genetic testing, as well as to become more centered on “underlying processes rather than clinical phenotypes,” the investigators report. They added that SUN criteria may present inefficiencies when individuals display more than 1 disorder, or experience a combination of infectious and non-infectious etiology. Also, the system performed better for intermediate uveitides because it included a classification for undifferentiated intermediate disease.

Interestingly, median treatment plus follow-up time for those who could be classified was 453 days, compared with 272 days in cases unable to be classified (P <.0001). This statistic may illustrate that SUN is designed to categorize based on initial presenting clinical features, but a number of uveitides progress, presenting added criteria with time.

Even with the large sample, there were many specific diagnoses with few cases each, which represents a limitation of the study. Also, the population at a single academic center may not be fully applicable to community clinics in ethnically-diverse locations. However, “in cases where classification by SUN could be attempted, the system performed well and generally agreed with their clinical diagnosis,” according to the analysis.

Reference

Mudie LI, Reddy AK, Patnaik JL, et al. Evaluation of the SUN classification criteria for the uveitides in an academic uveitis practice. Am Jour Ophthalmol. Published ahead of print on April 22, 2022. doi:10.1016/j.ajo.2022.04.007