PURPOSE
To develop imaging and consensus-based guidelines on the application of multimodal imaging in noninfectious multifocal choroiditis and panuveitis (MFCPU) and punctate inner choroiditis (PIC).
DESIGN
Consensus agreement guided by the review of literature and an expert committee using nominal group technique (NGT).
METHODS
An expert committee applied a timed structured nominal group technique (NGT) to achieve consensus-based recommendations on specific disease characteristics, biomarkers of activity, and complications for MFCPU and PIC. Representative cases with noninfectious active and inactive MFCPU and PIC with color fundus photographs (CFP), optical coherence tomography (OCT), fundus fluorescein angiography (FFA), OCT angiography (OCTA), indocyanine angiography (ICGA), and fundus autofluorescence images (FAF) were reviewed. These recommendations were voted upon by the entire task force.
RESULTS
The experts agreed that lesions of MFCPU and PIC can be well characterized using CFP. OCT is the preferred modality for detecting active lesions. Both FAF and OCT are effective for monitoring disease recurrence. Late-phase ICGA is most valuable in recurrent disease when the lesions are not visible on FAF and CFP. While OCTA and ICGA can successfully identify lesions and complications such as choroidal neovascularization, no imaging biomarkers were found to reliably distinguish between active and inactive lesions on these two modalities.
CONCLUSIONS
Incorporating imaging findings, particularly OCT, into the Standardization of Uveitis Nomenclature (SUN) classification criteria for MFCPU and PIC enables more precise assessment of disease activity. These consensus-based guidelines provide a framework for selecting optimal imaging modalities for diagnosis, monitoring and identification of complications of MFCPU and PIC.
M ultifocal choroiditis (MFC) is a nonspecific descriptive term used to encompass a variety of posterior uveitic conditions, that exhibit multifocal choroidal inflammation, with or without inflammation in other anatomic layers of the eye. Diseases included in this group may be infectious such as syphilis or tuberculosis, postinfectious such as presumed ocular histoplasmosis, inflammatory such as sarcoidosis, and isolated autoimmune or autoinflammatory entities such as punctate inner choroiditis (PIC) and multifocal choroiditis with panuveitis (MFCPU). Malignancies such as primary intraocular lymphoma and metastases can also masquerade as a MFC, complicating diagnosis and management.
The Standardization of Uveitis Nomenclature (SUN) defined MFCPU according to the following criteria: (1) absence of an associated infectious or systemic disease and (2) presence of multiple lesions larger than 125 µm in diameter predominantly located outside the vascular arcades. Lesions may manifest as inactive punched-out atrophic chorioretinal scars or may be associated with signs of intraocular inflammation, such as anterior chamber cells or vitreous inflammation.
Punctate inner choroiditis is a more specific term that is used to describe a noninfectious, nonsystemic, immune-mediated disease that targets the posterior pole. Whether PIC is on the same spectrum of MFCPU is disputed, but there is consensus that PIC lesions represent a more localized clinical phenotype compared to MFCPU. , As per the SUN classification criteria, PIC is defined according to the following criteria: (1) absence of appropriate infectious (eg, toxoplasmosis retinitis or syphilis) or associated systemic diseases (eg, sarcoidosis) and (2) multiple punctate lesions smaller than 250 µm in diameter, predominantly located in the posterior pole, with very little to no intraocular inflammation.
The SUN classification criteria for both MFCPU and PIC were developed following image review to characterize phenotypes but did not explicitly identify or define multimodal imaging biomarkers of activity specific to these conditions. This represents a gap in current clinical guidelines and emphasizes the need for more defined imaging characteristics to aid in the diagnosis, management, and monitoring of these uveitic conditions.
The M ultimodal Imaging in Uv eitis (MUV) task force is an international collaboration aimed at addressing this gap by developing imaging-based criteria to enhance the SUN classification framework. This effort is focused on five of the most common multifocal choroidopathies. This manuscript specifically focuses on imaging features of MFCPU and PIC.
METHODS
The Multimodal imaging in Uveitis (MUV) project is a research initiative of the International Uveitis Study Group (IUSG), an international group of uveitis specialists, to establish guidelines for the use of multimodal imaging in noninfectious posterior uveitis and to identify features characteristic of active and inactive disease for each entity. The study used previously collected, retrospectively reviewed, de-identified images and, as such, was considered “not human” research. The study was conducted under the tenets of the Declaration of Helsinki and received IRB exemption from the Vanderbilt University Medical Center (IRB # 240146).
SUBCOMMITTEE SELECTION
The MUV task force is comprised of uveitis specialists, retina specialists and ocular imaging experts with clinical and research experience in posterior uveitis. The MFCPU/PIC subcommittee, formed using purposeful sampling strategy, included geographically diverse specialists from India, Colombia, the Netherlands, and the United States to provide a broad range of expertise and perspectives, recognizing the regional variations in presentation and progression of MFCPU and PIC over time. This team was tasked with reviewing existing literature, analyzing multimodal image sets, and developing imaging guidelines and activity criteria with illustrative examples. We followed the principles of Standards for Reporting Qualitative Research: A Synthesis of Recommendations (SRQR). for reporting the results of our study.
CASE SELECTION
All task force members contributed deidentified high quality image sets that met SUN criteria , with clinical signs consistent with MFCPU and PIC, and exclusion of infectious causes. To be eligible for review, the image sets were required to include color fundus photographs (CFP), which were used as a surrogate for the clinical examination, optical coherence tomography (OCT), fundus fluorescein angiography (FFA), fundus autofluorescence (FAF), and indocyanine green angiography (ICGA) (if available) and OCT angiography (OCTA). No clinical information, patient identifiers or other patient demographics were shared to maintain confidentiality.
NOMINAL GROUP TECHNIQUE (NGT)
We employed a structured NGT approach, a formal consensus or brainstorming technique, in order to achieve agreement by the expert subcommittee. The NGT discussions were conducted virtually and led by one neutral facilitator (SG). The discussions ensured that time-limited, uninterrupted comments were provided by each committee member followed by anonymous voting. The propositions were either accepted (>75% super majority vote), defeated, or revised and revisited. The MFCPU/PIC subcommittee held multiple NGT discussions across time zones to ensure thorough review of all available image sets after excluding poor quality images. The subcommittee then determined the imaging biomarkers associated with active inflammation for each of the imaging modalities and developed consensus-based imaging guidelines in the diagnosis and management of MFCPU and PIC.
ESTABLISHMENT OF CONSENSUS
The subcommittee guidelines were shared with the entire MUV task force listed in Supplement A. The task force members used anonymous voting on an online survey platform and assessed the recommendations. Any requested modifications were collaboratively discussed among the members. The consensus was developed by the taskforce as follows:
Unanimous consensus: 100% participants agree
Strong consensus: > 95% vote
Consensus: 75% to 95% vote
Majority agreement: > 50% to 75% vote
No consensus: < 50% vote (lack of agreement or divided votes)
The percentage thresholds for consensus derived by voting were reported as per the guidelines of various international associations. These included the Guidelines International Network (GIN), European League Against Rheumatism (EULAR), and Association of Scientific Medical Associations of Germany (AWMF). When there was no consensus achieved (<50% vote), the guidelines were rejected.
Study data were collected and managed using REDCap electronic data capture tools hosted at Vanderbilt University Medical Center. , REDCap (Research Electronic Data Capture) is a secure, web-based software platform designed to support data capture for research studies, providing (1) an intuitive interface for validated data capture; (2) audit trails for tracking data manipulation and export procedures; (3) automated export procedures for seamless data downloads to common statistical packages; and (4) procedures for data integration and interoperability with external sources.
RESULTS
IMAGING FEATURES OF MULTIFOCAL CHOROIDITIS AND PUNCTATE INNER CHOROIDOPATHY
Color fundus photography
The subcommittee agreed with the SUN. , classification criteria that PIC lesions were smaller and located in the posterior pole. In contrast, MFCPU lesions were larger and could be identified in the posterior pole, nasal peripapillary retina, and in the periphery ( Figures 1 and 2 ). MFC and PIC are characterized by multifocal lesions which appear creamy and ill-defined when active. When inactive, the lesions are atrophic or punched out and clearly defined, often with a pigmented border. Peripapillary atrophy may be present but can also be detected in myopic eyes. Since myopia is associated with both MFC and PIC, as well as POHS, the subcommittee considered peripapillary atrophy to be a nonspecific finding.
Optical coherence tomography
The subcommittee unanimously agreed that OCT is the most reliable indicator of active inflammation in lesions associated with MFCPU and PIC. Active lesions demonstrated several characteristic features on OCT that the committee unanimously agreed represent classic indicators of disease activity. Fluffy subretinal hyper-reflective material (SHRM) ( Figure 3 ) overlying the retinal pigment epithelium (RPE) is a sign of active inflammation, with ellipsoid zone (EZ) disruption extending beyond the borders of the lesion ( Figure 4 ). There may be associated inflammatory pigment epithelial detachment (PED) or a discontinuous RPE band with elevated edges. Except for larger size of the MFCPU lesions, no differences are evident in the OCT appearance of MFCPU and PIC lesions. Figures 5 and 6 present serial follow-up of cases of MFCPU with SHRM indicating active disease that subsequently became inactive following treatment, with possible resolution of SHRM and EZ disruption. Enhanced-depth imaging OCT (EDI-OCT) shows focal thickening of the choroid underlying active lesions ( Figure 7 ). Inactive lesions are characterized by resolution of SHRM and reduction of the thickness of the inner choroid with clearer visualization of choroidal vessel landmarks. Focal disruption of EZ and RPE may remain ( Figure 8 ). Focal choroidal excavation may represent a longstanding complication of acute disease, best visualized on EDI-OCT.
The subcommittee did not reach consensus on the OCT appearance of chrysanthemum lesions or the pitchfork sign of MFCPU-associated choroidal neovascularization (CNV).
Fundus autofluorescence
On FAF, active lesions appear as new crops of uniformly hyper-autofluorescent (hyper-FAF) spots ( Figure 9 ). When the disease is reactivated, the new lesions are also hyper-FAF. Inactive lesions typically evolve to hypo-autofluorescent (hypo-FAF) lesions corresponding to atrophy; however, some inactive lesions can present with a hyper-FAF cuff ( Figure 10 ). The hyper-FAF cuff may be associated with subretinal fluid, scarring, or CNV. Such lesions can retain the hyper-FAF cuff and no longer indicate active disease. The areas of SHRM may resolve without RPE disruption, and hence may not necessarily be associated with hypo-FAF. The subcommittee agreed that ultra-wide field FAF is useful in determining lesion activity.
Fundus fluorescein angiography and optical coherence tomography angiography
The subcommittee determined that FFA did not offer any additional utility in the diagnosis or monitoring for reactivation of MFCPU and PIC lesions. Figure 11 shows leakage from the lesions when active. However, FFA is useful for detecting CNV which can display a hyperfluorescent lacy network with late leakage. Neovascular networks can also be captured with OCTA ( Figure 12 ). The subcommittee did not arrive at any consensus criteria for the activity of CNV on OCTA.
