Purpose
The purpose of this study was to determine classification criteria for multiple sclerosis-associated intermediate uveitis.
Design
Machine learning of cases with multiple sclerosis-associated intermediate uveitis and 4 other intermediate uveitides.
Methods
Cases of intermediate uveitides were collected in an informatics-designed preliminary database, and a final database was constructed of cases achieving supermajority agreement on the diagnosis, using formal consensus techniques. Cases were split into a training set and a validation set. Machine learning using multinomial logistic regression was used in the training set to determine a parsimonious set of criteria that minimized the misclassification rate among the intermediate uveitides. The resulting criteria were evaluated in the validation set.
Results
A total of 589 cases of intermediate uveitides, including 112 cases of multiple sclerosis-associated intermediate uveitis, were evaluated by machine learning. The overall accuracy for intermediate uveitides was 99.8% in the training set and 99.3% in the validation set (95% confidence interval: 96.1-99.9). Key criteria for multiple sclerosis-associated intermediate uveitis included unilateral or bilateral intermediate uveitis and multiple sclerosis diagnosed by the McDonald criteria. Key exclusions included syphilis and sarcoidosis. The misclassification rates for multiple sclerosis-associated intermediate uveitis were 0 % in the training set and 0% in the validation set.
Conclusions
The criteria for multiple sclerosis-associated intermediate uveitis had a low misclassification rate and appeared to perform sufficiently well enough for use in clinical and translational research.
M ultiple sclerosis is a neurologic disease characterized by demyelinating lesions in the brain or spinal column at 2 or more sites, occurring 2 or more times. , Typically, multiple sclerosis is a disease of young adults. Approximately 80% of cases present with a remitting and relapsing (remitting/relapsing) course and ∼20% with a primary progressive course. Patients presenting with remitting/relapsing multiple sclerosis typically have full recovery initially but may progress to relapse with persistent deficit and, ultimately, secondary progression. There is a strong environmental effect as the incidence and prevalence increase in populations farther away from the equator. , In Sub-Saharan Africa and East Asia, the prevalence of multiple sclerosis is estimated at 2.1-2.2/100,000 population, whereas in Canada it is estimated at 291/100,000 population. In the United States, the prevalence is estimated at 265-309/100,000 population. , Multiple sclerosis typically is diagnosed by using the McDonald criteria, which have been revised several times, most recently in 2017.
The most common ocular lesion appearing in multiple sclerosis is optic neuritis. Approximately 25% of patients with multiple sclerosis will present with optic neuritis, and as many as 70% will experience at least 1 episode of optic neuritis during their lifetime.
Patients with multiple sclerosis are reported to have an increased prevalence of uveitis. The reported prevalence of uveitis in patients with multiple sclerosis has ranged from 0.7%-28.6%, with the higher estimates from small case series and with an overall estimate of ∼1%. , These estimates are greater than the estimated prevalence of uveitis in the United States, which has been estimated at 69-114/100,000 population (approximately 0.1%). The reported prevalence of multiple sclerosis in series of patients with uveitis has ranged from 0.9%-3.1%, with an overall estimate of ∼1%, again, higher than the estimated prevalence of multiple sclerosis in the general population. However, interpretation of these data often has been hampered by “lumping” together all cases of uveitis or by anatomic “lumping,” making associations with specific types of uveitis difficult. Hence for many types of uveitis, it is uncertain whether the reported association is merely chance alone or a real statistical increase. Nevertheless, there appears to be a clear-cut association between multiple sclerosis and intermediate uveitis. The estimated prevalence of multiple sclerosis in intermediate uveitis has ranged from 2.3%-33% with an overall estimate of ∼11%, ∼10-fold higher than that in uveitis overall and ∼30-100-fold higher than that in the general population.
Intermediate uveitis refers to a class of uveitic diseases characterized by inflammation predominantly in the vitreous and an absence of retinitis and choroiditis. , Intermediate uveitides may be due to infections such as Lyme disease or syphilis; associated with systemic diseases, particularly sarcoidosis and multiple sclerosis; or may occur as an isolated, presumably immunity-mediated, ocular disorder of unknown cause. Eye-limited intermediate uveitis diagnoses include pars planitis characterized by snowball or snowbank formation, or both, and intermediate uveitis, non-pars planitis type, also known as undifferentiated intermediate uveitis.
Peripheral retinal vascular involvement is a characteristic feature of pars planitis and of multiple sclerosis-associated intermediate uveitis but is reported to be more common in multiple sclerosis-associated intermediate uveitis. It is typically asymptomatic and best appreciated on wide-field digital imaging, particularly fluorescein angiography. Angiographically, there may be venous leakage, staining, or occlusion, or both. Given the absence of differences among the multiple sclerosis disease features between multiple sclerosis patients with and without intermediate uveitis or peripheral retinal vascular changes, the pathogenetic significance of the association between peripheral retinal vascular changes and multiple sclerosis remains uncertain.
The Standardization of Uveitis Nomenclature (SUN) Working Group is an international collaboration which has developed classification criteria for 25 of the most common uveitides by using a formal approach to development and classification. , Among the intermediate uveitides studied was multiple sclerosis-associated intermediate uveitis.
METHODS
The SUN Developing Classification Criteria for the uveitides project proceeded in 4 phases as previously described: 1) informatics, 2) case collection, 3) case selection, and 4) machine learning. , , ,
Informatics
As previously described, the consensus-based informatics phase permitted the development of a standardized vocabulary and the development of a standardized, menu-driven hierarchical case collection instrument.
Case Collection and Case Selection
De-identified information was entered into the SUN preliminary database by the 76 contributing investigators for each disease as previously described. , , Cases in the preliminary database were reviewed by committees of 9 investigators for inclusion into the final database, using formal consensus techniques described in the accompanying articles. , Because the goal was to develop classification criteria, only cases with a supermajority agreement (>75%) that the case was the disease in question were retained in the final database (ie, were “selected”). ,
Machine Learning
The final database then was randomly separated into a training set (∼85% of the cases) and a validation set (∼15% of the cases) for each disease, as described in the accompanying article. Machine learning was used for the training set to determine criteria that minimized misclassification. The criteria then were tested in the validation set; for both the training set and the validation set, the misclassification rate was calculated for each disease. The misclassification rate was the proportion of cases classified incorrectly by the machine learning algorithm compared to the consensus diagnosis. For multiple sclerosis-associated, intermediate uveitis, the diseases against which it was evaluated were pars planitis; intermediate uveitis; non-pars planitis type; sarcoid intermediate uveitis; and syphilitic intermediate uveitis. Too few cases of Lyme disease-associated uveitis were collected in the data base for analysis by machine learning.
The study adhered to the principles of the Declaration of Helsinki. Institutional Review Boards (IRBs) at each participating center reviewed and approved the study; the study typically was considered either minimal risk or exempt by the individual IRBs.
RESULTS
A total of 183 cases of multiple sclerosis-associated intermediate uveitis were collected, and 112 cases (62%) achieved supermajority agreement for the diagnosis during the “selection” phase and were used in the machine learning phase. Those cases of multiple sclerosis-associated intermediate uveitis were compared to 477 cases of other intermediate uveitides, including 226 cases of pars planitis; 114 cases of intermediate uveitis, non-pars planitis type; 52 cases of sarcoidosis-associated intermediate uveitis; and 85 cases of syphilitic intermediate uveitis. The details of the machine learning results for those diseases are outlined in the accompanying article. The characteristics, at a presentation of a SUN Working Group investigator of cases with multiple sclerosis-associated intermediate uveitis type, are listed in Table 1 . The criteria developed after machine learning are listed in Table 2 . Key criteria were the presence of an intermediate uveitis and a diagnosis of multiple sclerosis. The 2017 McDonald criteria for the diagnosis of multiple sclerosis are outlined in Table 3 . The overall accuracy for intermediate uveitides was 99.8% in the training set and 99.3% in the validation set (95% confidence interval [CI]: 96.1-99.9). The misclassification rate for multiple sclerosis-associated intermediate uveitis in the training set was 0% and 0% in the validation set.
| Characteristic | Result |
|---|---|
| Number of cases | 112 |
| Demographics | |
| Median IQR (25th 75th percentile), y | 37 (30, 48) |
| Men, % | 15 |
| Women, % | 85 |
| Race/ethnicity, % | |
| White, non-Hispanic | 76 |
| Black, non-Hispanic | 4 |
| Hispanic | 2 |
| Asian, Pacific Islander | 1 |
| Other | 16 |
| Missing | 1 |
| Uveitis history | |
| Uveitis course, % | |
| Acute, monophasic | 3 |
| Acute, recurrent | 2 |
| Chronic | 85 |
| Indeterminate | 10 |
| Laterality, % | |
| Unilateral | 20 |
| Unilateral, alternating | 0 |
| Bilateral | 80 |
| Ophthalmic examination | |
| Keratic precipitates, % | |
| None | 74 |
| Fine | 10 |
| Round | 3 |
| Stellate | 2 |
| Mutton Fat | 5 |
| Other | 0 |
| Anterior chamber cells, % | |
| Grade 0 | 52 |
| ½+ | 21 |
| 1+ | 19 |
| 2+ | 9 |
| 3+ | 0 |
| 4+ | 0 |
| Hypopyon, % | 0 |
| Anterior chamber flare, % | |
| Grade 0 | 75 |
| 1+ | 21 |
| 2+ | 4 |
| 3+ | 0 |
| 4+ | 0 |
| Iris, % | |
| Normal | 82 |
| Posterior synechiae | 18 |
| Sectoral iris atrophy | 0 |
| Patchy iris atrophy | 0 |
| Diffuse iris atrophy | 0 |
| Heterochromia | 0 |
| IOP-involved eyes | |
| Median IQR (25th, 75th), mm Hg | 14 (12, 16) |
| Proportion of patients with IOP >24 mm Hg in either eye (%) | 1 |
| Vitreous cells, % a | |
| Grade 0 | 6 |
| ½+ | 24 |
| 1+ | 42 |
| 2+ | 25 |
| 3+ | 3 |
| 4+ | 0 |
| Vitreous haze, % a | |
| Grade 0 | 36 |
| ½+ | 28 |
| 1+ | 24 |
| 2+ | 11 |
| 3+ | 2 |
| 4+ | 0 |
| Vitreous snowballs | 54 |
| Pars plana snowbanks | 13 |
| Peripheral retinal vascular sheathing or leakage | 48 |
| Macular edema | 31 |
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
