Developing a Continuous Severity Scale for Macular Telangiectasia Type 2 Using Deep Learning and Implications for Disease Grading.

PURPOSE: Deep learning (DL) models have achieved state-of-the-art medical diagnosis classification accuracy. Current models are limited by discrete diagnosis labels, but could yield more information with diagnosis in a continuous scale. We developed a novel continuous severity scaling system for macular telangiectasia (MacTel) type 2 by combining a DL classification model with uniform manifold approximation and projection (UMAP). DESIGN: We used a DL network to learn a feature representation of MacTel severity from discrete severity labels and applied UMAP to embed this feature representation into 2 dimensions, thereby creating a continuous MacTel severity scale. PARTICIPANTS: A total of 2003 OCT volumes were analyzed from 1089 MacTel Project participants. METHODS: We trained a multiview DL classifier using multiple B-scans from OCT volumes to learn a previously published discrete 7-step MacTel severity scale. The classifiers' last feature layer was extracted as input for UMAP, which embedded these features into a continuous 2-dimensional manifold. The DL classifier was assessed in terms of test accuracy. Rank correlation for the continuous UMAP scale against the previously published scale was calculated. Additionally, the UMAP scale was assessed in the κ agreement against 5 clinical experts on 100 pairs of patient volumes. For each pair of patient volumes, clinical experts were asked to select the volume with more severe MacTel disease and to compare them against the UMAP scale. MAIN OUTCOME MEASURES: Classification accuracy for the DL classifier and κ agreement versus clinical experts for UMAP. RESULTS: The multiview DL classifier achieved top 1 accuracy of 63.3% (186/294) on held-out test OCT volumes. The UMAP metric showed a clear continuous gradation of MacTel severity with a Spearman rank correlation of 0.84 with the previously published scale. Furthermore, the continuous UMAP metric achieved κ agreements of 0.56 to 0.63 with 5 clinical experts, which was comparable with interobserver κ values. CONCLUSIONS: Our UMAP embedding generated a continuous MacTel severity scale, without requiring continuous training labels. This technique can be applied to other diseases and may lead to more accurate diagnosis, improved understanding of disease progression, and key imaging features for pathologic characteristics. FINANCIAL DISCLOSURE(S): Proprietary or commercial disclosure may be found in the Footnotes and Disclosures at the end of this article.

Training Deep Learning Models to Work on Multiple Devices by Cross-Domain Learning with No Additional Annotations.

PURPOSE: To create an unsupervised cross-domain segmentation algorithm for segmenting intraretinal fluid and retinal layers on normal and pathologic macular OCT images from different manufacturers and camera devices. DESIGN: We sought to use generative adversarial networks (GANs) to generalize a segmentation model trained on one OCT device to segment B-scans obtained from a different OCT device manufacturer in a fully unsupervised approach without labeled data from the latter manufacturer. PARTICIPANTS: A total of 732 OCT B-scans from 4 different OCT devices (Heidelberg Spectralis, Topcon 1000, Maestro2, and Zeiss Plex Elite 9000). METHODS: We developed an unsupervised GAN model, GANSeg, to segment 7 retinal layers and intraretinal fluid in Topcon 1000 OCT images (domain B) that had access only to labeled data on Heidelberg Spectralis images (domain A). GANSeg was unsupervised because it had access only to 110 Heidelberg labeled OCTs and 556 raw and unlabeled Topcon 1000 OCTs. To validate GANSeg segmentations, 3 masked graders manually segmented 60 OCTs from an external Topcon 1000 test dataset independently. To test the limits of GANSeg, graders also manually segmented 3 OCTs from Zeiss Plex Elite 9000 and Topcon Maestro2. A U-Net was trained on the same labeled Heidelberg images as baseline. The GANSeg repository with labeled annotations is at https://github.com/uw-biomedical-ml/ganseg. MAIN OUTCOME MEASURES: Dice scores comparing segmentation results from GANSeg and the U-Net model with the manual segmented images. RESULTS: Although GANSeg and U-Net achieved comparable Dice scores performance as human experts on the labeled Heidelberg test dataset, only GANSeg achieved comparable Dice scores with the best performance for the ganglion cell layer plus inner plexiform layer (90%; 95% confidence interval [CI], 68%-96%) and the worst performance for intraretinal fluid (58%; 95% CI, 18%-89%), which was statistically similar to human graders (79%; 95% CI, 43%-94%). GANSeg significantly outperformed the U-Net model. Moreover, GANSeg generalized to both Zeiss and Topcon Maestro2 swept-source OCT domains, which it had never encountered before. CONCLUSIONS: GANSeg enables the transfer of supervised deep learning algorithms across OCT devices without labeled data, thereby greatly expanding the applicability of deep learning algorithms.

CHARACTERISTICS OF PIGMENTED LESIONS IN TYPE 2 IDIOPATHIC MACULAR TELANGIECTASIA.

PURPOSE: Pigment in the midretina is a characteristic sign in Type 2 idiopathic macular telangiectasia (MacTel) and is considered to characterize the late stage of the disease. Our aim was to investigate its incidence, and relationship with risk factors for MacTel, including outer retinal vascularization and subretinal neovascular proliferation (SRNV). METHODS: Pigment extent was measured in fundus autofluorescence images of 150 eyes of 75 MacTel probands, using the Region Finder tool of Heidelberg Eye Explorer. A linear mixed model was used to analyze the dynamics of pigment and its associations with other features of the phenotype. The relative incidence of pigment and of outer retinal outer retinal vascularization and SRNV was analyzed within the full MacTel Study cohort (1,244 probands). RESULTS: Mean pigment area at baseline was 0.157 mm (range = 0-1.295 mm, SD = 0.228 mm, n = 101). Progression demonstrated a nonlinear pattern (P < 0.001) at an overall rate of 0.0177 mm/year and was associated with the initial plaque size and with SRNV. There was a strong correlation between fellow eyes (P ≤ 0.0001). In approximately 25% of all SRNV cases, SRNV may coincide with or precede pigment. CONCLUSION: Our data may be useful for refining the current system for staging disease severity in MacTel.

CORRELATION OF STRUCTURAL AND FUNCTIONAL OUTCOME MEASURES IN A PHASE ONE TRIAL OF CILIARY NEUROTROPHIC FACTOR IN TYPE 2 IDIOPATHIC MACULAR TELANGIECTASIA.

PURPOSE: Macular telangiectasia Type 2 is a bilateral, progressive, potentially blinding retinal disease characterized by both vascular and neurodegenerative signs. Both the area of the break in the ellipsoid zone seen in "en face" optical coherence tomographic (OCT) images and microperimetric focal retinal sensitivity loss have been proposed as potential measures of progression in macular telangiectasia. The authors aimed to assess the characteristics and interrelationship of these structural and functional disease markers from the data collected in a phase one clinical trial of ciliary neurotrophic factor in macular telangiectasia. METHODS: Orthogonal topographic (en face) maps of the ellipsoid zone were generated from Heidelberg Spectralis OCT volume scans (15 × 10° area, 30-µm B-scan intervals) or Zeiss Cirrus HD-OCT 4000 512 × 128 cube scans. Mesopic microperimetry was performed on CenterVue MAIA perimeters, using a Goldmann III stimulus in a custom test grid. Structural and functional data were analyzed by two methods: by calculating aggregate loss and by simple thresholding. The alignment quality of structural and functional data was also evaluated. RESULTS: Overall, the break area showed a good correlation with aggregate sensitivity loss (ρ = 0.834, P < 0.0001, 95% confidence interval 0.716-0.906) but also with the number of test points below a threshold value (e.g., <20 dB: ρ = 0.843, P < 0.0001, 95% confidence interval 0.755-0.902). Significant misalignment of the MAIA test grid was apparent in 13/48 visits of 7/14 eyes. CONCLUSION: The authors found a good correlation between ellipsoid zone break area and function loss. En face OCT mapping of the ellipsoid zone appears to demonstrate structural change before mesopic microperimetry can detect a focal loss of retinal sensitivity. Thresholding offers a quick alternative to calculating aggregate sensitivity loss.

SCOTOMA CHARACTERISTICS IN MACULAR TELANGIECTASIA TYPE 2: MacTel Project Report No. 7-The MacTel Research Group.

PURPOSE: To characterize scotomas in macular telangiectasia Type 2 (MacTel). METHODS: Five of the 27 centers performed microperimetry as part of the MacTel Natural History Observation Study. Data were analyzed in the Moorfields Eye Hospital Reading Centre. The number of stimuli under a threshold of 12, 10, 8, and <0 dB were counted (thresholding) and compared with one another. RESULTS: A total of 565 examinations were gradable, received from 632 eyes of 322 participants (age 61.1 ± 9.1 years, 62% females). The authors found absolute scotomas in 243 eyes (43%), 98% of these affected the temporal quadrant, and 99.5% were unifocal. Growth of absolute scotomas was limited to an extent of approximately 40 deg. Although transition from functionally unimpaired retina to absolute scotomas is generally steeply sloped, the larger a scotoma, the steeper it is. CONCLUSION: Scotoma features were consistent throughout a large MacTel cohort. The temporal quadrant was confirmed as predominantly affected, which might result from vascular or metabolic asymmetry. Functional loss did not exceed an area of 5° × 8° however advanced the disorder. Different MacTel phenotypes seem likely and point toward different types of progression; identifying these would improve planning for clinical trials and might lead to better understanding patient outcome.

CORRELATION OF CLINICAL AND STRUCTURAL PROGRESSION WITH VISUAL ACUITY LOSS IN MACULAR TELANGIECTASIA TYPE 2: MacTel Project Report No. 6-The MacTel Research Group.

PURPOSE: To evaluate progression of macular telangiectasia Type 2 lesions and their correlation with visual acuity. METHODS: An international multicenter prospective study with annual examinations including best-corrected visual acuity (BCVA), fundus photography, fluorescein angiography, and optical coherence tomography images graded centrally. Mixed models were used to estimate progression rates, and a generalized linear model to compute the relative risk of BCVA loss, loss of ellipsoid zone (EZ) reflectivity, development of pigment plaques, or neovascularization. RESULTS: One thousand and fourteen eyes of 507 participants were followed for 4.2 ± 1.6 years. Best-corrected visual acuity decreased 1.07 ± 0.05 letters (mean ± SE) per year. Of all eyes, 15% lost ≥15 letters after 5 years. Of the eyes without EZ loss, 76% developed a noncentral loss. Of the eyes with noncentral loss, 45% progressed to central EZ loss. The rate of BCVA loss in eyes with noncentral EZ loss at baseline was similar to eyes without EZ loss. The rate of BCVA loss was significantly higher in eyes with central EZ loss at baseline (-1.40 ± 0.14 letters, P < 0.001). CONCLUSION: Ellipsoid zone loss is frequently found in macular telangiectasia Type 2 and is an important structural component reflecting visual function. Its presence in the fovea significantly correlates with worse visual prognosis.

ABNORMAL RETINAL REFLECTIVITY TO SHORT-WAVELENGTH LIGHT IN TYPE 2 IDIOPATHIC MACULAR TELANGIECTASIA.

PURPOSE: Macular telangiectasia Type 2 (MacTel) is a bilateral, progressive, potentially blinding retinal disease characterized by vascular and neurodegenerative signs, including an increased parafoveal reflectivity to blue light. Our aim was to investigate the relationship of this sign with other signs of macular telangiectasia Type 2 in multiple imaging modalities. METHODS: Participants were selected from the MacTel Type 2 study, based on a confirmed diagnosis and the availability of images. The extent of signs in blue-light reflectance, fluorescein angiographic, optical coherence tomographic, and single- and dual-wavelength autofluorescence images were analyzed. RESULTS: A well-defined abnormality of the perifovea is demonstrated by dual-wavelength autofluorescence and blue-light reflectance in early disease. The agreement in area size of the abnormalities in dual-wavelength autofluorescence and in blue-light reflectance images was excellent: for right eyes: ρ = 0.917 (P < 0.0001, 95% confidence interval 0.855-0.954, n = 46) and for left eyes: ρ = 0.952 (P < 0.0001, 95% confidence interval 0.916-0.973, n = 49). Other changes are less extensive initially and expand later to occupy that area and do not extend beyond it. CONCLUSION: Our findings indicate that abnormal metabolic handling of luteal pigment and physical changes giving rise to increased reflectance are widespread in the macula throughout the natural history of the disease, precede other changes, and are relevant to early diagnosis.

Potential Effects of Hormone Therapy in Type 2 Idiopathic Macular Telangiectasia.

PURPOSE: To investigate the influence of hormone therapy with tamoxifen or estrogens on morphological changes in macular telangiectasia (MacTel) type 2 patients as revealed clinically in multiple imaging modalities. METHODS: Patients with a history of tamoxifen or estrogen use were selected from the cohort of the MacTel Study. A race-, age- and best-corrected visual acuity-matched group of MacTel participants not under hormone therapy served as the comparison group. The frequencies of typical features of the MacTel phenotype apparent in color fundus, red-free, fluorescein angiographic and optical coherence tomographic images were graded and analyzed statistically. RESULTS: Thirty-nine MacTel patients were included in the analyses, of whom 13 were receiving tamoxifen, 13 estrogens and 13 patients no hormone treatment. Patients treated with estrogens showed significantly fewer breaks in the ellipsoid zone on optical coherence tomography (7 eyes, 29.1%, vs. tamoxifen: 14 eyes, 53.8%, and vs. controls: 14 eyes, 53.8%, p = 0.04 in both analyses, Fisher exact test). Retinal crystalline deposits were significantly more frequent in patients receiving estrogens (12 eyes, 16.2%, vs. 2 eyes, 2.7%, p = 0.003, Fisher exact test). No significant between-group differences were apparent with regard to other features of the phenotype (extent of retinal low reflective spaces, late hyperfluorescence on fluorescein angiography or retinal thickness). CONCLUSIONS: Tamoxifen treatment does not seem to accentuate structural changes in patients with MacTel type 2. Treatment with estrogens may exhibit a neuroprotective effect as suggested by the decreased frequency of ellipsoid zone breaks in corresponding patients, although corroborative studies are warranted to confirm these exploratory data.

A Population-Based Ultra-Widefield Digital Image Grading Study for Age-Related Macular Degeneration-Like Lesions at the Peripheral Retina.

PURPOSE: Our understanding of the relevance of peripheral retinal abnormalities to disease in general and in age-related macular degeneration (AMD) in particular is limited by the lack of detailed peripheral imaging studies. The purpose of this study was to develop image grading protocols suited to ultra-widefield imaging (UWFI) in an aged population. DESIGN: A cross-sectional study of a random population sample in which UWFI was introduced at the 12-year review of the Reykjavik Eye Study in Iceland. PARTICIPANTS: Five hundred seventy-six subjects 62 years of age or older. METHODS: Ultra-widefield (up to 200°) color and autofluorescence images were obtained using the Optos P200CAF laser scanning ophthalmoscope (Optos plc, Dunfermline, Scotland). The images were graded at Moorfields Eye Hospital Reading Centre primarily based on the International Classification for AMD. Macular and peripheral changes were graded using a standardized grid developed for this imaging method. MAIN OUTCOME MEASURES: Presence or absence of hard, crystalline, and soft drusen; retinal pigment epithelial changes; choroidal neovascularization (CNV); atrophy; and hypoautofluorescence and hyperautofluorescence were graded in the peripheral retina. RESULTS: Of the eyes examined, 81.1% had AMD-like changes in the macula alone (13.6%), periphery alone (10.1%), and both periphery and macula (57.4%). There was no AMD-like CNV or pigment epithelial detachment in the periphery except in those cases in which these clearly originated from the macula. Seven patients had AMD-like atrophy in the periphery without end-stage disease in the macula. One patient with end-stage disease in the macula had normal periphery results on the color images. While analyzing the eyes, we detected pathologic appearances that were very reliably identified by graders. CONCLUSIONS: Phenotyping the retinal periphery using the categories defined by the International Classification confirmed the presence of wide-ranging AMD-like pathologic changes even in those without central sight-threatening macular disease. Based on our observations, we propose here new, reliably identifiable grading categories that may be more suited for population-based UWFI.

Multimodal imaging in type 2 idiopathic macular telangiectasia.

BACKGROUND: Macular telangiectasia Type 2 is a bilateral, progressive potentially blinding retinal disease characterized by both vascular and neurodegenerative signs that have been documented using different imaging techniques. The correlation between macular telangiectasia Type 2 signs from various imaging modalities is unknown. Our aim was to investigate the relationship of various macular telangiectasia Type 2 signs using fundus fluorescein angiography, optical coherence tomography and dual-wavelength autofluorescence images. METHODS: Participants were selected from the macular telangiectasia Type 2 Natural History Observation Study, based on a confirmed diagnosis and the availability of images. Signs in fundus fluorescein angiography, dual-wavelength autofluorescence, and optical coherence tomography images were graded according to standardized categories, and agreement among the multimodel imaging was assessed statistically. RESULTS: One hundred and ninety-one eyes of 96 patients were examined. Significant correlations were found between early and late fundus fluorescein angiography (ρ = 0.82, P < 0.0001), luteal pigment loss and early/late fundus fluorescein angiography signs (ρ = 0.52, P < 0.0001 and ρ = 0.62, P < 0.0001, respectively), inner and outer segment break length and pigment loss (Class 1 vs. 2/3, P < 0.0001; Class 2 vs. 3, P = 0.04). Correlation between pigment loss and retinal spaces/atrophic retinal restructuring was fair (κ = 0.25-0.33). Bilateral symmetry was slight to substantial (κ = 0.18-0.62). CONCLUSION: Our data demonstrate the relative extent of neurodegenerative and vascular signs; it may be useful for designing systems for staging disease severity using multimodal imaging and may also provide clues to the pathogenesis of the disease.

Associations between autofluorescence abnormalities and visual acuity in idiopathic macular telangiectasia type 2: MacTel project report number 5.

PURPOSE: The aim of this study was to determine whether typical abnormalities seen on autofluorescence (AF) imaging in patients with macular telangiectasia (MacTel) type 2 are correlated with visual acuity at presentation and with progression of visual loss over a 2-year follow-up period. METHODS: A subgroup of 218 patients (413 eyes) enrolled in the MacTel study that underwent AF imaging was included in the present study. Images were graded at the Moorfields Eye Hospital Reading Center. Recorded AF abnormalities at baseline and at 2 years included the presence of increased AF because of loss of masking at the central macula, localized decreased AF at the end of a retinal vessel, and large area of decreased AF. Best-corrected visual acuity was measured using the Early Treatment for Diabetic Retinopathy chart at baseline and after 2 years. Statistical associations were sought by means of a generalized linear model. RESULTS: Presence of increased macular AF (P = 0.004), a large area of decreased AF (P < 0.001), or decreased AF at the end of a retinal vessel (P < 0.001) at baseline were significantly associated with worse best-corrected visual acuity. Presence of increased macular AF (P < 0.001) or of localized decreased AF at the end of a retinal vessel (P < 0.001) and the absence of a large area of decreased AF (P < 0.001) were predictive of a subtle but significant drop in best-corrected visual acuity at 2 years. CONCLUSIONS: Increased central AF at baseline heralds worse best-corrected visual acuity and predicts further subtle visual loss in a period of 2 years, which, however, does not stand out from the overall slowly progressive natural history of the disease.

Self-Supervised Learning for Improved Optical Coherence Tomography Detection of Macular Telangiectasia Type 2.

Importance: Deep learning image analysis often depends on large, labeled datasets, which are difficult to obtain for rare diseases. Objective: To develop a self-supervised approach for automated classification of macular telangiectasia type 2 (MacTel) on optical coherence tomography (OCT) with limited labeled data. Design, Setting, and Participants: This was a retrospective comparative study. OCT images from May 2014 to May 2019 were collected by the Lowy Medical Research Institute, La Jolla, California, and the University of Washington, Seattle, from January 2016 to October 2022. Clinical diagnoses of patients with and without MacTel were confirmed by retina specialists. Data were analyzed from January to September 2023. Exposures: Two convolutional neural networks were pretrained using the Bootstrap Your Own Latent algorithm on unlabeled training data and fine-tuned with labeled training data to predict MacTel (self-supervised method). ResNet18 and ResNet50 models were also trained using all labeled data (supervised method). Main Outcomes and Measures: The ground truth yes vs no MacTel diagnosis is determined by retinal specialists based on spectral-domain OCT. The models' predictions were compared against human graders using accuracy, sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV), area under precision recall curve (AUPRC), and area under the receiver operating characteristic curve (AUROC). Uniform manifold approximation and projection was performed for dimension reduction and GradCAM visualizations for supervised and self-supervised methods. Results: A total of 2636 OCT scans from 780 patients with MacTel and 131 patients without MacTel were included from the MacTel Project (mean [SD] age, 60.8 [11.7] years; 63.8% female), and another 2564 from 1769 patients without MacTel from the University of Washington (mean [SD] age, 61.2 [18.1] years; 53.4% female). The self-supervised approach fine-tuned on 100% of the labeled training data with ResNet50 as the feature extractor performed the best, achieving an AUPRC of 0.971 (95% CI, 0.969-0.972), an AUROC of 0.970 (95% CI, 0.970-0.973), accuracy of 0.898%, sensitivity of 0.898, specificity of 0.949, PPV of 0.935, and NPV of 0.919. With only 419 OCT volumes (185 MacTel patients in 10% of labeled training dataset), the ResNet18 self-supervised model achieved comparable performance, with an AUPRC of 0.958 (95% CI, 0.957-0.960), an AUROC of 0.966 (95% CI, 0.964-0.967), and accuracy, sensitivity, specificity, PPV, and NPV of 90.2%, 0.884, 0.916, 0.896, and 0.906, respectively. The self-supervised models showed better agreement with the more experienced human expert graders. Conclusions and Relevance: The findings suggest that self-supervised learning may improve the accuracy of automated MacTel vs non-MacTel binary classification on OCT with limited labeled training data, and these approaches may be applicable to other rare diseases, although further research is warranted.

Translational PK/PD and the first-in-human dose selection of a PD1/IL15: an engineered recombinant targeted cytokine for cancer immunotherapy.

Introduction: Interleukin 15 (IL-15) is a potential anticancer agent and numerous engineered IL-15 agonists are currently under clinical investigation. Selective targeting of IL-15 to specific lymphocytes may enhance therapeutic effects while helping to minimize toxicities. Methods: We designed and built a heterodimeric targeted cytokine (TaCk) that consists of an anti-programmed cell death 1 receptor antibody (anti-PD-1) and an engineered IL-15. This "PD1/IL15" selectively delivers IL-15 signaling to lymphocytes expressing PD-1. We then investigated the pharmacokinetic (PK) and pharmacodynamic (PD) effects of PD1/IL15 TaCk on immune cell subsets in cynomolgus monkeys after single and repeat intravenous dose administrations. We used these results to determine the first-in-human (FIH) dose and dosing frequency for early clinical trials. Results: The PD1/IL15 TaCk exhibited a nonlinear multiphasic PK profile, while the untargeted isotype control TaCk, containing an anti-respiratory syncytial virus antibody (RSV/IL15), showed linear and dose proportional PK. The PD1/IL15 TaCk also displayed a considerably prolonged PK (half-life range ∼1.0-4.1 days) compared to wild-type IL-15 (half-life ∼1.1 h), which led to an enhanced cell expansion PD response. The PD was dose-dependent, durable, and selective for PD-1+ lymphocytes. Notably, the dose- and time-dependent PK was attributed to dynamic TMDD resulting from test article-induced lymphocyte expansion upon repeat administration. The recommended first-in-human (FIH) dose of PD1/IL15 TaCk is 0.003 mg/kg, determined based on a minimum anticipated biological effect level (MABEL) approach utilizing a combination of in vitro and preclinical in vivo data. Conclusion: This work provides insight into the complex PK/PD relationship of PD1/IL15 TaCk in monkeys and informs the recommended starting dose and dosing frequency selection to support clinical evaluation of this novel targeted cytokine.

Prevalence of age-related macular degeneration in Nakuru, Kenya: a cross-sectional population-based study.

BACKGROUND: Diseases of the posterior segment of the eye, including age-related macular degeneration (AMD), have recently been recognised as the leading or second leading cause of blindness in several African countries. However, prevalence of AMD alone has not been assessed. We hypothesized that AMD is an important cause of visual impairment among elderly people in Nakuru, Kenya, and therefore sought to assess the prevalence and predictors of AMD in a diverse adult Kenyan population. METHODS AND FINDINGS: In a population-based cross-sectional survey in the Nakuru District of Kenya, 100 clusters of 50 people 50 y of age or older were selected by probability-proportional-to-size sampling between 26 January 2007 and 11 November 2008. Households within clusters were selected through compact segment sampling. All participants underwent a standardised interview and comprehensive eye examination, including dilated slit lamp examination by an ophthalmologist and digital retinal photography. Images were graded for the presence and severity of AMD lesions following a modified version of the International Classification and Grading System for Age-Related Maculopathy. Comparison was made between slit lamp biomicroscopy (SLB) and photographic grading. Of 4,381 participants, fundus photographs were gradable for 3,304 persons (75.4%), and SLB was completed for 4,312 (98%). Early and late AMD prevalence were 11.2% and 1.2%, respectively, among participants graded on images. Prevalence of AMD by SLB was 6.7% and 0.7% for early and late AMD, respectively. SLB underdiagnosed AMD relative to photographic grading by a factor of 1.7. After controlling for age, women had a higher prevalence of early AMD than men (odds ratio 1.5; 95% CI, 1.2-1.9). Overall prevalence rose significantly with each decade of age. We estimate that, in Kenya, 283,900 to 362,800 people 50 y and older have early AMD and 25,200 to 50,500 have late AMD, based on population estimates in 2007. CONCLUSIONS: AMD is an important cause of visual impairment and blindness in Kenya. Greater availability of low vision services and ophthalmologist training in diagnosis and treatment of AMD would be appropriate next steps. Please see later in the article for the Editors' Summary.

Antibody-mediated coengagement of FcγRIIb and B cell receptor complex suppresses humoral immunity in systemic lupus erythematosus.

Engagement of the low-affinity Ab receptor FcγRIIb downregulates B cell activation, and its dysfunction is associated with autoimmunity in mice and humans. We engineered the Fc domain of an anti-human CD19 Ab to bind FcγRIIb with high affinity, promoting the coengagement of FcγRIIb with the BCR complex. This Ab (XmAb5871) stimulated phosphorylation of the ITIM of FcγRIIb and suppressed BCR-induced calcium mobilization, proliferation, and costimulatory molecule expression of human B cells from healthy volunteers and systemic lupus erythematosus (SLE) patients, as well as B cell proliferation induced by LPS, IL-4, or BAFF. XmAb5871 suppressed humoral immunity against tetanus toxoid and reduced serum IgM, IgG, and IgE levels in SCID mice engrafted with SLE or healthy human PBMC. XmAb5871 treatment also increased survival of mice engrafted with PBMC from a unique SLE patient. Unlike anti-CD20 Ab, coengagement of FcγRIIb and BCR complex did not promote B cell depletion in human PBMC cultures or in mice. Thus, amplification of the FcγRIIb inhibitory pathway in activated B cells may represent a novel B cell-targeted immunosuppressive therapeutic approach for SLE and other autoimmune diseases that should avoid the complications associated with B cell depletion.

Agreement between time-domain and spectral-domain optical coherence tomography in the assessment of macular thickness in patients with idiopathic macular telangiectasia type 2.

PURPOSE: To estimate the conversion factors to transpose macular thickness measurements on time-domain (TD) to various spectral-domain (SD) optical coherence tomography (OCT) machines in patients with macular telangiectasia type 2a (MacTel). PROCEDURES: Macular scans on TD- and SD-OCT were performed on patients at the same visit. The retinal thickness values in various ETDRS subfields and macular volume were compared between different OCT machines. RESULTS: The macular thickness and volume were significantly greater (p < 0.0001, r = 0.678-0.822) on SD-OCT. The mean differences in macular thickness between TD Stratus and SD-OCT by Spectralis, Cirrus and Topcon were 62, 41 and 20 µm, respectively. The conversion factor of macular thickness from TD-OCT to Spectralis, Cirrus and Topcon were +65, +39 and +25 µm, respectively. CONCLUSION AND MESSAGE: The estimates of conversion of macular thickness from TD- to SD-OCT using simple mean differences between machines and those by linear regression were similar suggesting that the former may be used for the longitudinal follow-up of MacTel patients.

Reduction of total IgE by targeted coengagement of IgE B-cell receptor and FcγRIIb with Fc-engineered antibody.

BACKGROUND: Sequestration of IgE to prevent its binding to high-affinity IgE receptor FcεRI on basophils and mast cells is an effective therapy for allergic asthma. IgE production requires differentiation of activated IgE(+) B cells into plasma cells upon allergen sensitization. B-cell receptor signaling is suppressed by the inhibitory IgG Fc receptor FcγRIIb; therefore, we reasoned that a therapeutic antibody that coengages FcγRIIb and IgE B-cell receptor would not only sequester IgE but also suppress its production by blocking IgE(+) B-cell activation and differentiation to IgE-secreting plasma cells. OBJECTIVE: To explore the effects of IgE sequestration versus IgE suppression by comparing omalizumab to FcγRIIb-optimized anti-IgE antibodies in humanized mouse models of immunoglobulin production. METHODS: By using a murine anti-IgE antibody as a template, we humanized, increased IgE binding, and modified its Fc domain to increase affinity for FcγRIIb. We next compared effects of this antibody (XmAb7195) versus omalizumab on the secretion of IgE and other isotypes in human PBMC cultures and in PBMC-engrafted severe combined immunodeficiency mice. RESULTS: Relative to omalizumab, XmAb7195 has a 5-fold higher affinity for human IgE and more than 400-fold higher affinity for FcγRIIb. In addition to sequestering soluble IgE, XmAb7195 inhibited plasma cell differentiation and consequent human IgE production through coengagement of IgE B-cell receptor with FcγRIIb. In PBMC-engrafted mice, XmAb7195 reduced total human IgE (but not IgG or IgM) levels by up to 40-fold relative to omalizumab. CONCLUSION: XmAb7195 acts by IgE sequestration coupled with an FcγRIIb-mediated inhibitory mechanism to suppress the formation of IgE-secreting plasma cells and reduce both free and total IgE levels.

Complex PK-PD of an engineered IL-15/IL-15Rα-Fc fusion protein in cynomolgus monkeys: QSP modeling of lymphocyte dynamics.

XmAb24306 is a lymphoproliferative interleukin (IL)-15/IL-15 receptor α (IL-15Rα) Fc-fusion protein currently under clinical investigation as an immunotherapeutic agent for cancer treatment. XmAb24306 contains mutations in IL-15 that attenuate its affinity to the heterodimeric IL-15 receptor ßγ (IL-15R). We observe substantially prolonged pharmacokinetics (PK) (half-life ∼ 2.5 to 4.5 days) in single- and repeat-dose cynomolgus monkey (cyno) studies compared to wild-type IL-15 (half-life ∼ 1 hour), leading to increased exposure and enhanced and durable expansion of NK cells, CD8+ T cells and CD4-CD8- (double negative [DN]) T cells. Drug clearance varied with dose level and time post-dose, and PK exposure decreased upon repeated dosing, which we attribute to increased target-mediated drug disposition (TMDD) resulting from drug-induced lymphocyte expansion (i.e., pharmacodynamic (PD)-enhanced TMDD). We developed a quantitative systems pharmacology (QSP) model to quantify the complex PKPD behaviors due to the interactions of XmAb24306 with multiple cell types (CD8+, CD4+, DN T cells, and NK cells) in the peripheral blood (PB) and lymphoid tissues. The model, which includes nonspecific drug clearance, binding to and TMDD by IL15R differentially expressed on lymphocyte subsets, and resultant lymphocyte margination/migration out of PB, expansion in lymphoid tissues, and redistribution to the blood, successfully describes the systemic PK and lymphocyte kinetics observed in the cyno studies. Results suggest that after 3 doses of every-two-week (Q2W) doses up to 70 days, the relative contributions of each elimination pathway to XmAb24306 clearance are: DN T cells > NK cells > CD8+ T cells > nonspecific clearance > CD4+ T cells. Modeling suggests that observed cellular expansion in blood results from the influx of cells expanded by the drug in lymphoid tissues. The model is used to predict lymphoid tissue expansion and to simulate PK-PD for different dose regimens. Thus, the model provides insight into the mechanisms underlying the observed PK-PD behavior of an engineered cytokine and can serve as a framework for the rapid integration and analysis of data that emerges from ongoing clinical studies in cancer patients as single-agent or given in combination.

Macular Telangiectasia Type 2: A Classification System Using MultiModal Imaging MacTel Project Report Number 10.

Purpose: To develop a severity classification for macular telangiectasia type 2 (MacTel) disease using multimodal imaging. Design: An algorithm was used on data from a prospective natural history study of MacTel for classification development. Subjects: A total of 1733 participants enrolled in an international natural history study of MacTel. Methods: The Classification and Regression Trees (CART), a predictive nonparametric algorithm used in machine learning, analyzed the features of the multimodal imaging important for the development of a classification, including reading center gradings of the following digital images: stereoscopic color and red-free fundus photographs, fluorescein angiographic images, fundus autofluorescence images, and spectral-domain (SD)-OCT images. Regression models that used least square method created a decision tree using features of the ocular images into different categories of disease severity. Main Outcome Measures: The primary target of interest for the algorithm development by CART was the change in best-corrected visual acuity (BCVA) at baseline for the right and left eyes. These analyses using the algorithm were repeated for the BCVA obtained at the last study visit of the natural history study for the right and left eyes. Results: The CART analyses demonstrated 3 important features from the multimodal imaging for the classification: OCT hyper-reflectivity, pigment, and ellipsoid zone loss. By combining these 3 features (as absent, present, noncentral involvement, and central involvement of the macula), a 7-step scale was created, ranging from excellent to poor visual acuity. At grade 0, 3 features are not present. At the most severe grade, pigment and exudative neovascularization are present. To further validate the classification, using the Generalized Estimating Equation regression models, analyses for the annual relative risk of progression over a period of 5 years for vision loss and for progression along the scale were performed. Conclusions: This analysis using the data from current imaging modalities in participants followed in the MacTel natural history study informed a classification for MacTel disease severity featuring variables from SD-OCT. This classification is designed to provide better communications to other clinicians, researchers, and patients. Financial Disclosures: Proprietary or commercial disclosure may be found after the references.

Prospective validation of a virtual clinic pathway in the management of choroidal naevi: the NAEVUS study Report no. 1: safety assessment.

BACKGROUND: Choroidal naevi are a common incidental finding prompting specialist referrals to ocular oncology. Rarely, such lesions have sufficient suspicious features to diagnose a small melanoma. The aim of the study is to show that 'virtual' imaging-based pathways are a safe and efficient option to manage such referrals. METHODS: A prospective cohort study at the Manchester Royal Eye Hospital and Moorfields Eye Hospital between June 2016 and July 2017 of the management decision of 400 patients reviewed by an ophthalmologist in a face-to-face consultation (gold standard) supported by fundus photography, optical coherence tomography, autofluorescence (AF) and B-mode ultrasound. The images were also read independently by blinded graders (non-medical) and blinded ophthalmologists, and a management decision was made based on image review alone (virtual pathway). The two pathways were compared for safety. RESULTS: The agreement for management decisions between face-to-face and virtual pathways was 83.1% (non-medical) and 82.6% (medical). There were more over-referrals in the virtual pathway (non-medical 24.3%, medical 23.3% of gold standard discharge) and only two under-referrals (10.5% of gold standard referrals), both borderline cases with minimal clinical risk. The agreement for risk factors of growth (orange pigment, subretinal fluid, hyper-AF) ranged between 82.3% and 97.3%. CONCLUSIONS: We prospectively validated a virtual clinic model for the safe management of choroidal naevi. Such a model of care is feasible with low rate of under-referral. An over-referral rate of almost 24% from the vitrual pathway needs to be factored into designing such pathways in conjunction with evidence on their cost-effectiveness.