Nonophthalmologist physicians do not confidently perform direct ophthalmoscopy. The use of artificial intelligence to detect papilledema and other optic-disk abnormalities from fundus photographs has not been well studied.We trained, validated, and externally tested a deep-learning system to classify optic disks as being normal or having papilledema or other abnormalities from 15,846 retrospectively collected ocular fundus photographs that had been obtained with pharmacologic pupillary dilation and various digital cameras in persons from multiple ethnic populations. Of these photographs, 14,341 from 19 sites in 11 countries were used for training and validation, and 1505 photographs from 5 other sites were used for external testing. Performance at classifying the optic-disk appearance was evaluated by calculating the area under the receiver-operating-characteristic curve (AUC), sensitivity, and specificity, as compared with a reference standard of clinical diagnoses by neuro-ophthalmologists.The training and validation data sets from 6779 patients included 14,341 photographs: 9156 of normal disks, 2148 of disks with papilledema, and 3037 of disks with other abnormalities. The percentage classified as being normal ranged across sites from 9.8 to 100%; the percentage classified as having papilledema ranged across sites from zero to 59.5%. In the validation set, the system discriminated disks with papilledema from normal disks and disks with nonpapilledema abnormalities with an AUC of 0.99 (95% confidence interval [CI], 0.98 to 0.99) and normal from abnormal disks with an AUC of 0.99 (95% CI, 0.99 to 0.99). In the external-testing data set of 1505 photographs, the system had an AUC for the detection of papilledema of 0.96 (95% CI, 0.95 to 0.97), a sensitivity of 96.4% (95% CI, 93.9 to 98.3), and a specificity of 84.7% (95% CI, 82.3 to 87.1).A deep-learning system using fundus photographs with pharmacologically dilated pupils differentiated among optic disks with papilledema, normal disks, and disks with nonpapilledema abnormalities. (Funded by the Singapore National Medical Research Council and the SingHealth Duke-NUS Ophthalmology and Visual Sciences Academic Clinical Program.).
Obstructive sleep apnea (OSA) is a common pulmonary disorder with many systemic sequelae. Its association with diseases seen in neuro-ophthalmology is being increasingly recognized. Treatment of OSA with continuous positive airway pressure (CPAP) may reverse the pathology in some eye diseases.We conducted a search of the literature using the Ovid Medline database and Google Scholar, focusing on articles that explored the association between OSA and ophthalmic conditions. These included hypertensive retinopathy, diabetic retinopathy, central serous retinopathy (CSR), retinal vein occlusion (RVO), nonarteritic anterior ischemic optic neuropathy (NAION), glaucoma, idiopathic intracranial hypertension (IIH), papilledema, and stroke.There is evidence of an independent association between OSA, and both hypertension and Type 2 diabetes mellitus, as well as the retinopathy associated with each disease. Evidence also suggests a link between OSA and CSR, NAION, and stroke. The evidence remains controversial or insufficient for convincing causative association with RVO, glaucoma, and IIH. However, in patients presenting with CSR, and IIH, underlying OSA is a common finding. CPAP reverses some pathological changes, but the evidence is limited.OSA is associated with many diseases seen in neuro-ophthalmology clinics. More studies are required to assess the real ability of CPAP to reverse pathological changes. Ophthalmologists can screen for undiagnosed OSA in patients presenting with certain eye diseases.
Myelin oligodendrocyte glycoprotein (MOG-IgG) antibodies have been associated with a variety of demyelinating neurologic disorders, including optic neuritis. It remains unclear whether the presence of MOG-IgG represents a distinct syndrome or is simply a marker for central demyelination. Two experts, John J. Chen, MD, PhD, and Clare L. Fraser, MBBS, MMed, debate this topic. Pro: John J. Chen, MD, PhD Opening Statement In medicine, physicians and other providers often adopt the preference of being either a "lumper" or "splitter" when it comes to disease processes with overlapping characteristics. However, once a molecular basis is identified that reliably distinguishes one disease entity from another, particularly in a situation for which such a distinction affects treatment, being a "lumper" could lead to delays in our diagnosis, initiation of best therapies, and understanding of the disease process. The understanding of neuromyelitis optica spectrum disorder (NMOSD) is a perfect example of how serologic diagnosis has advanced characterization of demyelinating disease. Just over a decade ago, there was debate as to whether neuromyelitis optica (NMO) was a separate entity from multiple sclerosis (MS) (1). However, the discovery of antibodies to aquaporin-4 (AQP4-IgG) in 2004 cemented NMOSD as a separate disorder and revolutionized our understanding of the pathophysiology, clinical characteristics, and treatment of the disease (2,3). More recently, antibodies against myelin oligodendrocyte glycoprotein (MOG-IgG) have emerged as a reproducible marker for a subset of patients with optic neuritis and other demyelinating event phenotypes. While there is some clinical overlap with other demyelinating disorders, MOG-IgG–associated demyelinating disease is now becoming recognized as its own disease entity that is distinct from classic MS and AQP4-IgG–positive NMOSD (4,5). Characteristics suggestive of myelin oligodendrocyte glycoprotein disease The most common phenotype of MOG-IgG–positive demyelinating disease is optic neuritis, particularly when recurrent, followed by myelitis, acute disseminating encephalomyelitis (ADEM), and brainstem encephalitis (4–9). There are some characteristics of MOG-IgG–positive demyelinating disease that should alert the clinician to this possibility. Compared to other forms of acute demyelinating optic neuritis, MOG-IgG–positive optic neuritis has a higher likelihood of being recurrent, bilateral, and associated with prominent disc edema. Recurrent optic neuritis is seen in between 50% and 80% of cases of MOG-IgG–positive optic neuritis (6,7,9). This condition can sometimes be steroid responsive and dependent, thus meeting the criteria for what has previously been termed chronic relapsing inflammatory optic neuropathy (9–11). Bilateral simultaneous involvement occurs in almost 50% of cases of MOG-IgG–positive optic neuritis (6,7,9,12,13). Optic disc edema at onset is present in up to 86% (4,9,12–15). The disc edema can be severe, with peripapillary hemorrhages; these are a feature that is rarely seen in other forms of demyelinating optic neuritis. The vision loss is usually severe at the nadir, but recovery is typically better than that seen with AQP4-IgG–positive optic neuritis (7,9). On MRI, there is often longitudinally extensive enhancement of the optic nerve in those with MOG-IgG–positive optic neuritis (4,9,14,16,17). Perineural enhancement of the optic nerve sheath and peribulbar structures is seen in up to 50% of cases and is a fairly specific sign of this disorder, which is not typically seen with MS or AQP4-IgG–positive optic neuritis (9,15,16,18,19). Patients with longitudinally extensive transverse myelitis (≥3 contiguous vertebral segments) and negative AQP4-IgG antibody testing should be evaluated for MOG-IgG status; this is the case because such extensive spinal cord involvement is rarely seen in patients with classic MS. Transverse myelitis involves the conus medullaris in MOG-IgG disease more commonly than in other demyelinating diseases (including AQP4-IgG–positive transverse myelitis) (4,7,20). A recent study also found that the T2-signal abnormalities in the spinal cord are often restricted to the grey matter, forming a hallmark "H-sign" on axial images (20). An accompanying brainstem encephalitis and/or an ADEM-like presentation should also raise suspicion for MOG-IgG disease. These phenomena are less commonly seen in patients with AQP4-IgG–positive disease or in those with classic MS. Myelin oligodendrocyte glycoprotein–positive demyelinating disease is distinct from multiple sclerosis Although antibodies to myelin oligodendrocyte glycoprotein (MOG) were initially associated with MS based on nonspecific solid phase assay results (21), recent studies using transfected cell-based assays have found that MOG antibodies are almost never seen in patients with typical MS. In the process of optimizing the MOG-IgG assay at the Mayo Clinic, 50 patients with classic MS were tested, and none of them were positive for MOG-IgG (22). A follow-up study evaluating 86 patients with MOG-IgG–positive optic neuritis found only 1 patient with MS; this patient had a minimally elevated MOG-IgG binding index of 2.8 (laboratory cutoff of 2.5). In addition, none of the patients in that study had oligoclonal bands in the cerebral spinal fluid (CSF) (9). Many other groups have also found that patients with MOG-IgG–associated demyelinating disease do not have oligoclonal bands in the CSF and do not follow a classic MS disease course (5,7). Finally, a multicenter study of 200 patients and review of the literature found only 1 single borderline-positive MOG-IgG result among 290 patients with MS (23). The lack of coexisting MOG antibodies in patients with classic MS indicates that MOG-IgG disease is a distinct and separate process. Myelin oligodendrocyte glycoprotein–positive demyelinating disease is distinct from AQP4-IgG–positive neuromyelitis optica spectrum disorder Patients with antibodies to MOG can develop optic neuritis and longitudinally extensive transverse myelitis and thus fulfill the criteria for NMOSD, the disease process classically associated with AQP4-IgG. Approximately 30% of patients with NMOSD are seronegative for AQP4-IgG; recent studies have suggested that MOG-IgG is positive in approximately one-third of these patients (5,24,25). Much like MOG antibodies are rare in patients with MS, MOG-IgG is almost never seen in patients who have antibodies to AQP4 (26). This supports the concept that MOG-IgG disease and AQP4-IgG–positive NMOSD are separate entities (5). While there is clinical overlap between MOG-IgG–mediated and AQP4-IgG–mediated disease, the pathophysiologies are very different. Pathologic specimens from patients with AQP4-IgG–positive NMOSD show astrocyte destruction and secondary demyelination (27). In contrast, MOG-IgG–positive inflammatory disease tissue shows primary demyelination with preserved astrocytes; this was previously designated as Pattern II demyelination (5,26,28,29). Therefore, AQP4-IgG and MOG-IgG appear to be fundamentally distinct entities with different underlying pathophysiological bases. The importance of testing for myelin oligodendrocyte glycoprotein and recognizing myelin oligodendrocyte glycoprotein–positive demyelinating disease as a separate entity Testing for MOG antibodies, and recognizing MOG-IgG–positive demyelinating disease as a separate entity, truly matter because these distinctions influence our diagnostic ability, prognostication, and ultimately, our treatment of the patient. MOG-IgG disease can present with widespread central nervous system (CNS) inflammation that can be concerning for a vasculitic or infectious process. Much of what we know about the pathology for MOG-IgG disease was derived from brain biopsies performed because of diagnostic uncertainty; these were obtained before the advent of reliable cell-based assays for MOG-IgG. Now that we have a better understanding of the phenotype of MOG-IgG disease and have specific assays for MOG antibodies, diagnosing MOG-IgG disease with a simple serum test can lead to the correct diagnosis and exclude the necessity of a brain biopsy. As our understanding of MOG-IgG disease improves, formal diagnostic criteria will be developed and further refined in order to reliably diagnose this disorder. In addition, separating MOG-IgG disease from other demyelinating diseases will improve our understanding of the natural course. This will enhance our abilities to prognosticate and counsel our patients. Presumably because of the differences in pathogenesis, MOG-IgG–associated inflammation has better outcomes than AQP4-IgG disease even among patients who meet the current criteria for NMOSD. Despite a tendency to cause recurrent severe optic neuritis, the majority of patients with MOG-IgG–positive optic neuritis have meaningful recovery of vision and retain functional vision (7,9,30). This is unlike patients with AQP4-IgG–positive optic neuritis, for whom over one-third have poor visual outcomes (31–33). Recognizing MOG-IgG–demyelinating disease not only is helpful in diagnosis and prognosis but also may have a substantial impact on treatment. Recent studies have shown that MS disease-modifying agents are not effective in preventing relapses in the setting of MOG-IgG–associated disease (18,34,35). Treatment with MS disease-modifying agents could lead to the accumulation of CNS lesion burden from continued relapses; this could unwittingly lead to addition or escalation of what are actually ineffective disease-modifying agents, thus subjecting patients to unnecessary side effects. In addition, it is possible that MS disease-modifying agents could even worsen MOG-IgG–associated disease; this phenomenon has been seen in the setting of AQP4-IgG–positive NMOSD (4,5,18,36–38). The optimal treatments for AQP4-IgG and MOG-IgG diseases may be different as well. A recent multicenter study suggested that rituximab reduces relapse rates in MOG-IgG disease but not as effectively as it does for patients AQP4-IgG–positive NMOSD (39). Therefore, in the future, it will be important for patients with demyelinating disease to be classified according to their underlying molecular diagnosis rather than by a set of clinical criteria alone. Appreciating MOG-IgG–associated disease as its own entity will allow us to better understand the disease pathogenesis. This will be important because it will ultimately lead to directed therapies. Such a paradigm is being tested in ongoing clinical trials for AQP4-IgG–positive NMOSD. Lumping MOG-IgG–positive disease with other forms of demyelinating processes will hamper advancements that can be made for this unique entity. Con: Clare L. Fraser, MBBS, MMed, FRANZCO MOG is expressed exclusively in the CNS as a minor component of myelin. The protein structure of MOG is classified as an immunoglobulin and is found preferentially at the extracellular surface; MOG thus serves as a marker of oligodendrocyte maturation (40). MOG is also thought to serve in myelin adhesion, integrity, and cellular interactions (41). It is therefore studied as a target in CNS demyelinating disease. The potential role for MOG-IgG antibodies in AQP4-negative NMOSD was first suggested in 2007 (6). Following this, in vitro and patient cohort studies have pursued this link. Taking this line of thought one step further, it is suggested that MOG-IgG positivity may denote a disease entity in its own right. However, some of the studies have been limited by the assay type used, small patient numbers, limited diversity of the patients reviewed, and the lack of long-term follow-up data. Identification of MOG-IgG antibodies using cell-based assays (transfected or transduced with native human MOG in its conformational state and analyzed by flow cytometry or microscopy) have demonstrated the presence of this antibody in pediatric patients with ADEM or a relapsing demyelinating (MS-like) disease. Further studies using cell-based assays have shown MOG-IgG positivity in patients with NMOSD. Therefore, the data must be reviewed carefully before we decide if MOG antibody–associated optic neuritis is indeed a distinct entity. Consequence, not cause? To place MOG antibodies in the context of the current clinical literature, and thus this debate, it is important to review the animal studies. Experimental autoimmune encephalomyelitis (EAE) is an animal model of CNS demyelination. Induction of EAE requires immunizing the animal with CNS tissue homogenates or purified myelin components (42). This results in a complex immune response, including a strong T-cell–driven component. Transfer of encephalitogenic T cells can also initiate demyelination and EAE in animals. While MOG antibodies are part of this response, they alone do not necessarily result in the transfer of disease and are not required for severe clinical disease (40,43). This implies that immune system exposure to neurological tissue may result in MOG-IgG as a secondary consequence of preexisting damage; this is similar to the way in which antiretinal antibodies are found in conditions like retinitis pigmentosa. When the mouse IgG monoclonal antibody (mAb) equivalent of anti-MOG, known as 8-18C5 mAb, was transferred into animals that already had EAE, a hyperacute inflammatory response and extensive demyelinating plaques were seen. This suggests that perhaps MOG antibodies amplifies and modifies preexisting demyelinating pathology (44). Furthermore, this effect was dependent on the T cells having weakened the blood–brain barrier; there was no correlation between the antibody titers and the clinical disease (43). In mice, MOG-IgG only causes temporary damage of myelin and axons. More importantly, it does not produce inflammatory cell infiltration, axonal loss, neural degeneration, or astrocyte death (45). It could therefore be argued that perhaps MOG antibodies amplify preexisting disease rather than being a direct and separate pathological entity. The early published clinical research also points toward MOG antibodies being a broader consequence of neurological disease. In 1991, one group reported MOG-IgG in the CSF of 7 patients with MS, in 2 patients with "other inflammatory neurological disease" (OIND), and in 1 patient with tension headaches (46). Larger studies found MOG-IgG in 14%–33% of MS patients, 19%–55% of OIND, and in 3%–8% of noninflammatory neurological disease patients; this included all cases of neurosarcoid tested (47,48). MOG-IgG was also found in 10% of rheumatoid arthritis patients who had no neurological disease (47). Studies from this era are limited by the use of enzyme-linked immunosorbent assay (ELISA), which is less reliable than the newer assays. Even the newer cell-based assays are not without problems. In one study that used full-length human MOG, 48% of epilepsy control patients had a positive test result for MOG-IgG, which reduced to 5.8% when an IgG1-specific secondary antibody was used (49). These results would argue that MOG-IgG is a more generalized marker of inflammation, rather than a disease-causing antibody, in many forms of neurological disease. Chronic inflammatory CNS disease may induce autoantibodies by virtue of epitope spreading. In the MS literature, one study of 103 patients with clinically isolated syndrome found that 21% of patients were positive for both MOG-IgG and myelin basic protein IgM antibodies; 41% were positive for MOG antibodies alone. Those with antibodies to one or both myelin components were more likely to have relapses and to meet the criteria for clinically definite MS (21). The authors went on to emphasize that they could not prove whether the measured antibodies had demyelinating capacity or whether they represented an epiphenomenon of myelin destruction. Postmortem studies also showed higher levels of MOG antibodies within the MS lesions, compared to CSF and serum, suggesting local production as a consequence of disease (50). The authors also reported similar tissue findings in a patient with CNS aspergillosis. While most studies since this time have found MOG-IgG exclusively in patients with optic neuritis and/or myelitis who are AQP4-IgG negative, some concerns have been raised about the data (18). The cohorts included a median of 9 patients with only 24-month median follow-up; long-term follow-up was not available (6). Some cohorts contained no Caucasian patients or were genetically mixed. This may be of relevance as genetic (HLA-DRB-1 types) and infectious (Chlamydia pneumoniae, Helicobacter pylori) factors are thought to contribute to the pathogenesis of NMOSD (51). Finally, some control cohorts were too small to assess the specificity of the tests (6). Lumper or splitter? NMOSD is a demyelinating disorder of the CNS, typically presenting with optic neuritis or transverse myelitis (3). Since the discovery of AQP4 antibodies, it is now distinguishable from MS. However, 10%–25% of clinical NMOSD patients are negative for the AQP4 antibody. Therefore, assuming that MOG-IgG seropositivity does not per se constitute an "alternative diagnosis" from MS, then MOG-related conditions could still fit the 2015 diagnostic criteria for NMOSD (3). Jarius et al (18) found that 32% of patients with MOG-IgG met the 2015 NMOSD criteria, while 44% fulfilled the McDonald criteria for MS at the time of their study. However, AQP4-IgG–positive NMOSD is a disease of astrocytes, whereas MOG-IgG targets oligodendrocytes and therefore might be classified as a form of opticospinal MS (52). Perhaps, there are several routes to characterizing the final common pathways of the diseases we know as MS and NMOSD. Indeed, NMOSD may have 3 subtypes: AQP4-IgG–positive, MOG-IgG–positive, and dual positive cases. In a study of 174 patients, 2 cases tested positive to both MOG-IgG and AQP4-IgG antibodies (53). These 2 patients were women in their 50s who presented with bilateral simultaneous optic neuritis and longitudinally extensive transverse myelitis. Both patients tested positive for MOG-IgG and AQP4-IgG in both the serum and CSF. Mader et al (54) found that one-third of MOG-IgG–positive patients who fulfilled the diagnostic criteria for NMOSD also were AQP4-IgG positive. Dual serum positivity has also been reported in 1 patient with isolated optic neuritis in Japan and in 1 patient in China (55,56). In another study of recurrent optic neuritis (2 episodes separated by more than 1 month), 6 of 23 patients tested positive for both antibodies (57). Of these 6 patients, 50% failed to respond to high-dose corticosteroids and plasmapheresis, with visual acuity remaining poor. Using fluorescence-activated cell sorting, one group reported 10 patients (8%) with dual positivity to MOG-IgG and AQP4-IgG from a group of 125 patients with NMOSD (58). The majority of these patients have MS-like brain lesions on MRI, severe edematous multifocal changes on spine MRI, and pronounced loss of retinal nerve fiber layer thickness on optical coherence tomography, even in clinically unaffected eyes. The disease was typically multiphasic, with a high annual relapse rate and severe residual disability by Expanded Disability Status Scale and visual acuity testing. Yan et al (58) argue that a lack of similar findings in other studies was consequence of laboratory techniques that only allowed for the detection of antibodies in the extracellular or cell-surface domains. To date, I have not found any reports of double-positive antiacetylcholine receptor and muscle-specific kinase antibodies antibodies in myasthenia gravis, yet both forms of the disease are called myasthenia gravis! Why create 2 separate entities for AQP4 antibody and MOG antibody–positive demyelinating disease? Finally, the term MOG-IgG optic neuritis may be a misnomer because 80%–93% of such patients develop a relapsing disease; in addition, 52% develop more widespread neuroinflammatory changes, including myelitis, brainstem encephalitis and cerebellitis (6). Therefore, it seems more appropriate to say MOG-IgG–associated disease (new acronym—MAD?) spectrum disorder rather than MOG-IgG optic neuritis. Rebuttal: John J. Chen, MD, PhD Dr. Fraser has brought up several points that advance discussion of the importance of MOG antibodies in demyelinating disease. Dr. Fraser mentioned that prior studies on MOG-IgG were limited by small patient numbers. However, there are now many recent reports from large cohorts of patients spanning multiple ethnicities, with studies being published in Asia, Australia, the United Kingdom, France, the United States, Brazil, and many other countries (6,7,9,12,13,59,60). These larger studies have provided great insight into MOG-IgG–positive disease and have further shown that this is a distinct entity. I agree with Dr. Fraser that the early studies on MOG-IgG were fraught with difficulties and a lack of specificity; this incorrectly led to the notion that MOG-IgG was associated with MS. However, as Dr. Fraser mentioned, studies in this era were limited by the use of ELISA, which did not evaluate antibodies to MOG in its native form, leading to the poor specificity (61). The new cell-based assays use MOG in its native form which, in conjunction with optimization of the secondary antibodies, have led to very good specificity. Antibodies to MOG are not found in normal patients, those with classic MS, patients with other optic neuropathies (glaucoma cohort), or in those with other autoimmune disease (lupus) (22,23,49). While there are rare cases of simultaneous MOG and AQP4 antibody positivity in the same patient, these are exceedingly rare with the use of the new cell-based assays and are limited to case reports. In all of the large recent series published on MOG-IgG, there have not been any cases of dual positivity for both MOG-IgG and AQP4-IgG (6,7,9,13). While Dr. Fraser reported several series of patients with coinciding MOG-IgG and AQP4-IgG, these studies were older and utilized less-specific assays that were either not cell based or not optimized with respect to the secondary antibodies. There are likely rare cases of double positivity because both MOG-IgG and AQP4-IgG demyelinating disease are autoimmune disorders, but this is an exception rather than the rule. Therefore, autoantibodies to MOG are specific for a unique subset of patients with demyelinating disease and are not seen in other disease processes. In addition, they are not just an epiphenomenon of inflammatory CNS or optic nerve disease. It is still unclear whether MOG antibodies are directly pathogenic or merely a marker of disease. A recent study demonstrated that MOG antibodies derived from humans with MOG-IgG disease were pathogenic and, in fact, induced demyelination on intrathecal injection in 2 different rat EAE models (62). Weber et al (5) summarized other potential pathogenic mechanisms that are distinct from AQP4-IgG–mediated disease. Even if antibodies to MOG end up not being directly pathogenic, it is clear that MOG-IgG is a good marker of a specific disease process that is distinct from AQP4-IgG–mediated disease and from MS. This distinction has tremendous clinical ramifications and cannot be ignored because of the implications for prognosis and treatment. While some cases of MOG-IgG disease will meet the current criteria for NMOSD or the McDonald criteria for MS, these reflect the lack of specificity in the diagnostic criteria rather than of MOG-IgG disease not being its own entity. There are unique elements to MOG-IgG disease that distinguish it from its MS and AQP4-IgG counterparts. As discussed above, MOG-IgG disease has a different presentation and prognosis. MOG-IgG disease is seen equally in males and females, unlike MS and AQP4-IgG disease that both have a female predilection. The CSF cytokine profile for MOG-IgG–positive disease is distinct from that of MS (63). As discussed before, the pathology found on brain biopsy for MOG-IgG is entirely different than what is found in AQP4-IgG disease. The best treatment for MOG-IgG has yet to be established, but it is clear that disease-modifying agents used to treat MS are not effective. Therefore, MOG-IgG–positive disease has a different pathogenesis, prognosis, and treatment. Lumping MOG-IgG together into the same disease process will hamper the understanding of this unique demyelinating process. Similar to how NMO became distinct from MS over a decade ago with the discovery of the novel biomarker AQP4-IgG, now that we have a specific and reliable marker for MOG-IgG disease, it will also become its own separate disease. I agree with Dr. Fraser that MOG-IgG optic neuritis is not the most appropriate name because MOG-IgG–positive disease can have a diverse neuroinflammatory phenotype. This has been recently called MOG-IgG encephalomyelitis according to a group of international experts (4). While the name will likely change, its relevance as a distinct disease will not. Recognizing MOG-IgG disease as its own distinct entity will allow us to better understand the disease process and improve treatments for this potentially debilitating disease. Rebuttal: Clare L. Fraser, MBBS, MMed, FRANZCO Over the past few years, the weight of evidence is leaning toward MOG-IgG optic neuritis being a distinct entity, which Dr. Chen has nicely summarized. I am glad that we agree that it does seem more accurate to give a broader title to the disease entity, such as MOG-IgG encephalomyelitis, rather limiting the condition to MOG-IgG optic neuritis only; this is true particularly in children. Some of my own arguments in Part 1 were drawn from the old literature, with less accurate methods for antibody testing than we now have available. In 2016, one of the early articles showed that patients with MOG antibody–associated demyelination appeared to have a unique clinical, radiological, and therapeutic profile (64). Over the past few years, more clinical experience has led me to view and manage MOG-IgG–positive patients as having a separate disease process to MS and NMO. Our Australian experience of MOG-IgG–associated demyelination was recently published (12). The clinical course, therapeutic response, and outcomes of 59 patients were described. The article emphasized that there remains diversity in phenotypes associated with MOG-IgG–associated demyelination and that some overlap may be present between patients with clinically definite MS and MOG antibodies. However, given the emerging literature on the clinical and radiological phenotype and pattern of treatment response, I agree with Dr. Chen that this condition warrants designation as a separate clinical entity from MS and NMOSD. In adults, the article recommended that MOG-IgG antibody testing could be reasonably restricted to patients with a clinical and radiological phenotype atypical for MS, particularly in the event of isolated or recurrent optic neuritis (12). While MOG-IgG–associated demyelination manifests as an optic neuritis in the majority of adult patients, there is a wide clinical spectrum that appears to be broader than simply AQP4-IgG–negative patients with NMOSD. Given the high pretest probability of MOG-IgG positivity in children, it seems reasonable to test for MOG antibodies in all childhood-onset demyelination, particularly if relapsing (12). Conclusions: Andrew G. Lee, MD, and Gregory Van Stavern, MD The evidence to date suggests that MOG-IgG–associated demyelinating disease may represent a distinct disorder with clinical, radiologic, and pathologic features that distinguish it from both MS and NMOSD. Whether one is a lumper or a splitter, however, the clinical spectrum of MOG is still expanding. It remains to be seen whether MOG antibodies are a marker for immune-mediated demyelination or whether these are intrinsically pathogenic, per se. This may have implications for treatment and prognosis for MOG-IgG–associated demyelinating disease. Further studies and clinical experience during the next several years will increase our understanding of this important condition.
Ophthalmic manifestations of allergic fungal sinusitis (AFS) are rare, but can occur in advanced disease. A 32-year-old man with advanced AFS presented with severe bilateral vision loss and restricted ocular motility. Magnetic resonance imaging and histological analysis confirmed active chronic AFS. Functional endoscopic sinus surgery was performed, with adjunctive steroid therapy. Although AFS is a reasonably well-recognised entity, severe disease causing bilateral visual deficits is rarely encountered. This can confound the diagnosis and appropriate treatment. Ophthalmologists should thus be aware of compressive optic neuropathy as a complication of advanced AFS to prompt early treatment and mitigate visual loss.
In recent years, CTLA-4 and PD-1/PD-L1 checkpoint inhibitors have proven to be effective and have become increasingly popular treatment options for metastatic melanoma and other cancers. These agents work by enhancing autologous antitumor immune responses. Immune-related ophthalmologic complications have been reported in association with checkpoint inhibitor use but remain incompletely characterized. This study seeks to investigate and further characterize the neuro-ophthalmic and ocular complications of immune checkpoint blockade treatment.A survey was distributed through the secure electronic data collection tool REDCap to neuro-ophthalmology specialists in the North American Neuro-Ophthalmology Society listserv. The study received human subjects approval through the University of California at Los Angeles Institutional Review Board. The survey identified patients sent for neuro-ophthalmic consultation while receiving one or more of a PD-1 inhibitor (pembrolizumab, nivolumab, or cemiplimab); PD-L1 inhibitor (atezolizumab, avelumab, or durvalumab); or the CTLA-4 inhibitor ipilimumab. Thirty-one patients from 14 institutions were identified. Patient demographics, neuro-ophthalmic diagnosis, diagnostic testing, severity, treatment, clinical response, checkpoint inhibitor drug used, and cancer diagnosis was obtained.The checkpoint inhibitors used in these patients included pembrolizumab (12/31), nivolumab (6/31), combined ipilimumab with nivolumab (7/31, one of whom also received pembrolizumab during their course of treatment), durvalumab (3/31), ipilimumab (2/31), and cemiplimab (1/31). Malignant melanoma (16/31) or nonsmall cell lung carcinoma (6/31) were the most common malignancies. The median time between first drug administration and the time of ophthalmological symptom onset was 14.5 weeks. Eleven patients had involvement of the optic nerve, 7 patients had inflammatory orbital or extraocular muscle involvement, 6 patients had ocular involvement from neuromuscular junction dysfunction, 4 patients had cranial nerve palsy, and 4 patients had non neuro-ophthalmic complications. Use of systemic corticosteroids with or without stopping the checkpoint inhibitor resulted in improvement of most patients with optic neuropathy, and variable improvement for the other ophthalmic conditions.This study describes the variable neuro-ophthalmic adverse events associated with use of immune checkpoint inhibitors and contributes a more thorough understanding of their clinical presentations and treatment outcomes. We expect this will increase awareness of these drug complications and guide specialists in the care of these patients.
'/%3e%3c/defs%3e%3cpath fill='%23012169' d='M0 0h10080v5040H0z'/%3e%3cpath stroke='%23fff' stroke-width='.6' d='m0 0 6 3m0-3L0 3' clip-path='url(%23b)' transform='scale(840)'/%3e%3cpath stroke='%23e4002b' stroke-width='.4' d='m0 0 6 3m0-3L0 3' clip-path='url(%23c)' transform='scale(840)'/%3e%3cpath stroke='%23fff' stroke-width='840' d='M2520 0v2520M0 1260h5040'/%3e%3cpath stroke='%23e4002b' stroke-width='504' d='M2520 0v2520M0 1260h5040'/%3e%3cg fill='%23fff'%3e%3cuse xlink:href='%23d' x='2520' y='3780'/%3e%3cuse xlink:href='%23a' x='7560' y='4200'/%3e%3cuse xlink:href='%23a' x='6300' y='2205'/%3e%3cuse xlink:href='%23a' x='7560' y='840'/%3e%3cuse xlink:href='%23a' x='8680' y='1869'/%3e%3cuse xlink:href='%23e' x='8064' y='2730'/%3e%3c/g%3e%3c/svg%3e)
'%3e%3cpath fill='%23012169' d='M0 0v30h60V0z'/%3e%3cpath stroke='%23fff' stroke-width='6' d='m0 0 60 30m0-30L0 30'/%3e%3cpath stroke='%23C8102E' stroke-width='4' d='m0 0 60 30m0-30L0 30' clip-path='url(%23b)'/%3e%3cpath stroke='%23fff' stroke-width='10' d='M30 0v30M0 15h60'/%3e%3cpath stroke='%23C8102E' stroke-width='6' d='M30 0v30M0 15h60'/%3e%3c/g%3e%3c/svg%3e)
