To determine classification criteria for multifocal choroiditis with panuveitis (MFCPU).
Machine learning of cases with MFCPU and 8 other posterior uveitides.
Cases of posterior uveitides were collected in an informatics-designed preliminary database, and a final database was constructed of cases achieving supermajority agreement on diagnosis, using formal consensus techniques. Cases were split into a training set and a validation set. Machine learning using multinomial logistic regression was used on the training set to determine a parsimonious set of criteria that minimized the misclassification rate among the posterior uveitides. The resulting criteria were evaluated on the validation set.
One thousand sixty-eight cases of posterior uveitides, including 138 cases of MFCPU, were evaluated by machine learning. Key criteria for MFCPU included (1) multifocal choroiditis with the predominant lesions size >125 µm in diameter; (2) lesions outside the posterior pole (with or without posterior involvement); and either (3) punched-out atrophic chorioretinal scars or (4) more than minimal mild anterior chamber and/or vitreous inflammation. Overall accuracy for posterior uveitides was 93.9% in the training set and 98.0% (95% confidence interval 94.3, 99.3) in the validation set. The misclassification rates for MFCPU were 15% in the training set and 0% in the validation set.
The criteria for MFCPU had a reasonably low misclassification rate and seemed to perform sufficiently well for use in clinical and translational research.
I n 1984 Dreyer and Gass described a new posterior uveitic disease, multifocal choroiditis with panuveitis (MFCPU). The disease had a retinal picture similar to that of the presumed ocular histoplasmosis syndrome in that there were “punched-out atrophic” chorioretinal scars of variable size, but differed in that there was a variable anterior chamber and vitreous inflammation and there was no evidence of prior histoplasmosis infection on serologic testing, skin testing, and chest radiography. Most cases were bilateral. There often were lesions of variable character, including “active lesions” described as yellow to yellow-white, round or oval, sometimes irregular, and mildly elevated, with “punched-out atrophic” chorioretinal scars with variable hyperpigmentation at the edges. Lesions typically were >250 µm in size. ,
Multifocal choroiditis with panuveitis is an uncommon uveitic disease. Most data on the disease come from case series. In 1 6-year series of all patients with uveitis seen at a single, tertiary-case uveitis center in Australia, MFCPU accounted for 2.4% of all uveitic cases. The incidence of MFCPU has been estimated at 0.03 cases per 100,000 population per year. Multifocal choroiditis with panuveitis has been reported with a wide age range, but most cases occurred in young to middle-aged adults. Although MFPCU occurs in both men and women, the majority of reported cases have been in women. It has been reported in multiple ethnic groups, but the majority of cases seem to occur in white individuals. –
The etiology of MFCPU is unknown, and it is unassociated with a systemic disease. The differential diagnosis of MFCPU includes those diseases that can produce a multifocal choroidopathy, such as punctate inner choroiditis (PIC), syphilis, tuberculosis in endemic areas, and sarcoidosis. Rarely, late-stage, untreated birdshot chorioretinitis (BSCR) may have a similar appearance. Occasionally serpiginous choroiditis not adjacent to the disc may look like MFCPU, but the characteristic fluorescein angiogram of serpiginous choroiditis typically allows the correct diagnosis to be made. ,
The active lesions of MFCPU have been described as yellow-orange, round or oval, sometimes elevated, and typically >250 µm in size. The “punched-out” atrophic scars involve loss of choroid and retinal pigment epithelium in a circular fashion, typically with pigment clumping at the edge. Fluorescein angiography of active MFCPU lesions has been reported as demonstrating multiple chorioretinal spots with progressive hyperfluorescence throughout the angiogram. By contrast, the atrophic scars demonstrate window defects on fluorescein angiography. Occasionally it can be difficult to differentiate choroidal neovascularization from an active MFCPU lesion of fluorescein angiography alone. On indocyanine green angiography, active MFCPU lesions have been reported to present as hypofluorescent spots that fade by the late phases of the angiogram, suggesting that the lesions are at the level of the choriocapillaris and/or retinal pigment epithelium. Fundus autofluorescence imaging has been reported as useful in assessing the activity of MFCPU lesions. Atrophic scars are hypoautofluorescent, whereas active lesions are mildly hyperautofluorescent. More lesions may be visible on fundus autofluorescence imaging than are evident clinically. Optical coherence tomography is useful in diagnosing macular edema and active choroidal neovascularization. It also has been reported to distinguish active choroidal lesions from atrophic scars. Optical coherence tomography angiography, although not routinely used, has been reported to differentiate choroidal neovascularization from active MFCPU lesions by detecting the abnormal subretinal vessels.
Reported structural complications include macular edema, choroidal neovascularization, optic neuropathy, epiretinal membranes, and cataract. , , , , Choroidal neovascularization has been reported as the most common cause of vision loss. Incidences of visual impairment (20/50 or worse) and blindness (20/200 or worse) in involved eyes have been estimated at 0.19/eye-year (EY) and 0.12/EY, respectively, and in the better-seeing eye at 0.07/EY and 0.04/EY, respectively. High-dose oral corticosteroids have been reported to control the inflammation and decrease the occurrence of retinal structural complications, but doses low enough for long-term use (<10 mg/day) appear to be largely ineffective. , Conversely, immunosuppression has been reported to reduce the occurrence of structural complications by more than 80%. , Hence, if treatment is needed, oral corticosteroids and immunosuppression appear to be the preferred approach. , Choroidal neovascularization typically is treated with adjunctive anti–vascular endothelial growth factor therapy.
The Standardization of Uveitis Nomenclature (SUN) Working Group is an international collaboration that has developed classification criteria for 25 of the most common uveitides using a formal approach to development and classification. Among the diseases studied was MFCPU.
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.
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 selection into the final database, using formal consensus techniques described in the accompanying article. , 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”). ,
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 on the training set to determine criteria that minimized misclassification. The criteria then were tested on 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 when compared to the consensus diagnosis. For MFCPU, the diseases against which it was evaluated were acute posterior multifocal placoid pigment epitheliopathy, BSCR, multiple evanescent white dot syndrome, PIC, serpiginous choroiditis, sarcoidosis-associated posterior uveitis, syphilitic posterior uveitis, and tubercular uveitis.
The study adhered to the principles of the Declaration of Helsinki. Institutional review boards at each participating center reviewed and approved the study; the study typically was considered either minimal risk or exempt by the individual institutional review boards.
Two hundred fifty-one cases of MFCPU were collected, and 138 (57%) achieved supermajority agreement on the diagnosis during the “selection” phase and were used in the machine learning phase. These cases of MFCPU were compared to cases of posterior uveitides, including 82 cases of acute posterior multifocal placoid pigment epitheliopathy, 207 cases of BSCR, 51 cases of multiple evanescent white dot syndrome, 122 cases of serpiginous choroiditis, 144 cases of PIC, 12 cases of sarcoid posterior uveitis, 35 cases of syphilitic posterior uveitis, and 277 cases of tubercular posterior uveitis or panuveitis uveitis. The details of the machine learning results for these diseases are outlined in the accompanying article. The characteristics of cases with MFCPU are listed in Table 1 , and the classification criteria developed after machine learning are listed in Table 2 . Key features of the criteria include multifocal choroiditis with round or oval lesions >125 µm in size, involvement of the midperiphery and/or periphery, and punched-out atrophic scars ( Figure 1 ) or active lesions with more than minimal vitritis. The overall accuracies for posterior uveitides were 93.9% in the training set and 98.0% (95% confidence interval 94.3, 99.3) in the validation set. The misclassification rate for MFCPU in the training set was 15%, and in the validation set 0%. The diseases with which MFCPU most often was confused in the training set were BSCR and PIC.
|Number of cases||138|
|Age, median, years (25th, 75th percentile)||38 (28, 55)|
|Asian, Pacific Islander||3|
|Uveitis course (%)|
|Keratic precipitates (%)|
|Anterior chamber cells, grade (%)|
|Anterior chamber flare, grade (%)|
|Iris atrophy (sectoral, patchy, or diffuse)||0|
|IOP, involved eyes|
|Median, mm Hg (25th, 75th percentile)||16 (13, 18)|
|Proportion of patients with IOP > 24 mm Hg either eye (%)||4|
|Vitreous cells, grade (%)|
|Vitreous haze, grade (%)|
|Chorioretinal lesion characteristics|
|Lesion number (%)|
|Unifocal (1 lesion)||0|
|Paucifocal (2-4 lesions)||5|
|Multifocal (≥5 lesions)||89|
|Lesion shape & character (%)|
|Ameboid or serpentine||0|
|Oval or round||94|
|Inflammatory lesion/scar location (%) a|
|Posterior pole only involved||1.5|
|Posterior pole and periphery/midperiphery||56.5|
|Midperiphery and periphery only||42|
|Typical lesion size (%)|
|Other features (%)|
|Retinal vascular sheathing||9|
|Retinal vascular leakage||13|