Madhura A. Tamhankar

Objectives: To validate claims-based algorithms for identifying thyroid eye disease (TED) cases and assess whether machine learning can improve case identification in a large, linked real-world dataset. Methods: Using a large, linked database from Komodo Health^® and Academy IRIS^® Registry, we evaluated six rule-based algorithms incorporating Graves’ disease (GD), eye symptoms and signs. The IRIS Registry’s curated data, based on confirmed TED diagnoses from medical notes, served as the reference standard. Additionally, we developed supervised machine learning models using demographic, diagnostic, procedural, and medication data. Feature selection was performed using recursive feature elimination to rank predictive codes and construct a simplified, interpretable model. Cross-validation was used to assess model performance and compare performance with the rule-based algorithms. Results: The rule-based algorithms demonstrated a trade-off between sensitivity and specificity, with some achieving high specificity but limited sensitivity. Algorithm 1 had the highest sensitivity (48.7%) but lower specificity (59.9%) and PPV (75.8%). Algorithms 2–5 demonstrated higher specificity (87.2–93.5%) but lower sensitivity (17.8–27.0%). Algorithm 6 improved sensitivity (33.4%) compared to Algorithms 2–5 while maintaining high specificity (86.8%) and a strong PPV (86.7%). Machine learning models demonstrated similar trade-offs. One model achieved improved specificity (77.2%) with sensitivity of 49.3%, outperforming Algorithm 1 in specificity while matching its sensitivity. Another model maximized specificity (91.7%) and PPV (89.8%) at a reduced sensitivity of 28.5%. These results highlight the flexibility of machine learning models in adjusting performance to address different research objectives. Conclusions: This study evaluated existing rule-based algorithms for identifying TED cases in claims data, revealing trade-offs between sensitivity and specificity. Machine learning models provide additional flexibility, allowing performance to be tailored to specific research use cases. While no single method consistently outperformed others across all metrics, both rule-based and machine learning approaches demonstrated value in improving TED case identification using real-world data sources.

Background: To characterize the retreatment course of patients with thyroid eye disease (TED), who had reactivation after initial therapy with teprotumumab. Methods: This was a single-center longitudinal cohort study of patients who received an initial course of teprotumumab for active TED and were followed for at least 6 months. Reactivation was defined as the increase of proptosis of 2 mm or more or an increase in Clinical Activity Score (CAS) of two points or more, as adapted from the Optic-X study. Data collection included patient age, sex, smoking status, history of thyroidectomy or radioactive iodine, proptosis measurements, clinical activity score (CAS) before initial infusion of teprotumumab, time interval to reactivation, diplopia assessment by the Bahn–Gorman scale, CAS at the time of reactivation, and CAS and proptosis measurements after completion of retreatment and retreatment modalities, including clinical monitoring, corticosteroids, teprotumumab, and/or surgery. Among the reactivated cohort, the treatment response of patients who received a second course of teprotumumab was compared with patients who were treated with intravenous (IV) steroids. Results: Twenty-six percent (11/42) of patients experienced reactivation of TED with an average time to reactivation of 9 (SD:5) months (range: 2–20 months), average CAS at reactivation of 4 (SD:1) (range: 3–7), and average increase in proptosis of 3 (SD:1) mm (range: 2–6 mm). Of the 11 patients who reactivated, 4 received a second course of teprotumumab, while 6 received IV steroids. One patient elected to monitor. The patients who received a second course of teprotumumab had a mean (SD) posttreatment CAS score of 0 reduction in proptosis of 4 (2) mm (range: 3–6). The patients who received IV steroids had a mean (SD) posttreatment CAS of 2 (1) (range: 1–4) and a reduction in proptosis of 0 (1) mm (range: [−1] to [2]). Univariate analyses to look at predictors of reactivation found no correlation between factors such as age, sex, duration of TED, smoking status, presence of diplopia, previous treatment with radioactive iodine, history of periorbital surgery, and/or thyroidectomy after initial completion of teprotumumab between the 2 cohorts. We found a significant correlation between the CAS scores before initial treatment ( P = 0.036) and thyroid hormone dysregulation ( P = 0.006) in those who experienced reactivation. Conclusions: Patients with TED may experience reactivation of the disease after initial therapy with teprotumumab. Reactivated disease responds to repeat therapy with teprotumumab with higher previous CAS and thyroid hormonal dysregulation being the variables that were significantly associated with reactivation. These data underscore the importance of long-term monitoring and exploring underlying triggers for disease reactivation. Understanding these factors could help predict which patients may require retreatment or chronic dosing with teprotumumab. Further studies are essential to advance our understanding of the immunomodulatory effects of teprotumumab, duration of its therapeutic benefit, and potential retreatment strategies to improve long-term patient outcomes.

Background/Objectives: Stroke damage to the primary visual cortex induces large, homonymous visual field defects that impair daily living. Here, we asked if vision-related quality of life (V-QoL) is impacted by time since stroke. Subjects/Methods: We conducted a retrospective meta-analysis of 95 occipital stroke patients (F/M=26/69, 27-78 years old, 0.5 to 373.5 months post-stroke) in whom V-QoL was estimated using the National Eye Institute Visual Functioning Questionnaire (NEI-VFQ) and its 10-item neuro-ophthalmic supplement (Neuro10). Visual deficit severity was represented by the Perimetric Mean Deviation (PMD) calculated from 24-2 Humphrey visual fields. Data were compared with published cohorts of visually-intact controls. The relationship between V-QoL and time post-stroke was assessed across participants, adjusting for deficit severity and age with a multiple linear regression analysis. Results: Occipital stroke patients had significantly lower NEI-VFQ and Neuro10 composite scores than controls. All subscale scores describing specific aspects of visual ability and functioning were impaired except for ocular pain and general health, which did not differ significantly from controls. Surprisingly, visual deficit severity was not correlated with either composite score, both of which increased with time post-stroke, even when adjusting for PMD and age. Conclusions: V-QoL appears to improve with time post-occipital stroke, irrespective of visual deficit size or patient age at insult. This may reflect the natural development of compensatory strategies and lifestyle adjustments. Thus, future studies examining the impact of rehabilitation on daily living in this patient population should consider the possibility that their V-QoL may change gradually over time, even without therapeutic intervention.

Background: Teprotumumab, an insulin-like growth factor I receptor inhibitory antibody, improved proptosis, diplopia, inflammatory signs/symptoms, and quality of life in patients with active thyroid eye disease (TED) in clinical trials. The trials excluded patients with dysthyroid optic neuropathy (DON). Recently, many case reports and case series have reported the successful use of teprotumumab to treat DON. Here, we review the data from published cases and our clinical experience in treating patients having DON with teprotumumab. Methods: A literature search was conducted of patients with DON treated with teprotumumab from January 2020 through September 2022. Data from DON patients from the authors' (M.A.T. and C.A.B.) clinical practice were included. Primary outcome measure was mean (SD) improvements for visual acuity, color vision, and visual fields. Improvements in proptosis and clinical activity score (CAS) and diplopia were compared before and after teprotumumab administration. Results: Ten observational studies/case reports were identified along with 2 patients in our practice. In all, there were 24 active TED patients with DON (37 eyes) who were treated with teprotumumab. Mean (SD) age was 66.5 (13.6) years and 13 (54%) were females, disease duration ranged from 2 months to >15 years. 22/24 patients had none, minimal improvement or progression of visual loss with intravenous/oral corticosteroids, orbital decompression (n = 9), and orbital radiation (n = 2). There were 2 patients who received teprotumumab as the only therapy. Overall, 88% (21/24) reported improvement in visual acuity after teprotumumab and in 75% (18/24), improvement in vision was observed after just 2 infusions of teprotumumab. Three eyes had decompression surgery in close proximity to teprotumumab infusions and were excluded from analyses. Mean (SD) improvement in visual acuity was 3.73 lines (SD 3.74), range 2–15 lines in 33 eyes. The mean (SD) improvement in the mean deviation on visual field testing in 15 eyes was 5.6 db (3.0 db). Mean (SD) improvement in proptosis was 4.37 mm (SD: 2.11) (20 patients, 32 eyes); and clinical activity score: mean reduction of 5.1 (1.3) for 18 patients. Teprotumumab was well tolerated in all but one patient. Adverse events reported included fatigue, dysgeusia, hearing loss, nausea, hyperglycemia, and muscle spasms. Conclusions: Teprotumumab is an effective treatment for DON in our experience and in published cases in whom treatment with steroids, surgery, or orbital radiation was unsuccessful.

Background: To identify the frequency and etiologies of visual disturbances after cataract surgery in patients referred to Neuro-ophthalmology. Methods: This study is a retrospective chart review. Records of patients 18 years and older referred to neuro-ophthalmology clinics for new-onset visual disturbances within 6 months of cataract surgery were reviewed. Those with pre-existing neuro-ophthalmic disorders, combined intraocular procedures with cataract surgery, or inadequate follow-up were excluded. The main outcome measures were frequency and etiologies of visual disturbances after cataract surgery. Secondary analyses of a cohort of patients who had cataract surgery at our institution were performed to determine the frequency and etiology of visual disturbances after uneventful cataract surgery. Results: One hundred seventy-three patients met the inclusion criteria (internal referral: 36/173, from outside surgeons: 137/173). Sixty-one percent (106/173) were newly diagnosed with neuro-ophthalmic etiologies, including 21% (36/173) with afferent and 40% (70/173) with efferent disorders. Thirty-six percent (62/173) of patients had non neuro-ophthalmic causes and 3% (5/173) had systemic conditions responsible for visual disturbances postoperatively. Decompensated strabismus causing diplopia was the most common neuro-ophthalmic diagnosis after cataract surgery (50%, 53/106). Of the 13,715 patients who had cataract surgery performed at our institution over a 9-year period, 20 of 36 patients referred for visual disturbances were identified with neuro-ophthalmic etiologies of which 85% (17/20) had postoperative diplopia. Conclusions: In our study, decompensated strabismus causing diplopia was the most common neuro-ophthalmic visual disturbance after cataract surgery. Detailed history and ocular alignment should be assessed before cataract surgery to identify patients with the risk.

Background: Chronic inflammatory demyelinating polyneuropathy (CIDP) is an acquired, immune-mediated, and clinically heterogeneous demyelinating disease affecting the nerve roots and peripheral nerves. We report a series of 4 patients who presented with early and progressive vision loss in the context of new-onset CIDP: 3 due to papilledema and 1 due to optic neuropathy without papilledema. Methods: This was a retrospective case series of 4 patients with vision loss as a presenting feature of CIDP evaluated at the Hospital of the University of Pennsylvania from January 2016 to August 2021. Demographic, clinical, diagnostic, and treatment data were collected via retrospective medical record review. Results: Case 1 was a 51-year-old man with 2 months of progressive bilateral papilledema associated with reduced visual acuity (count fingers at 1 foot in each eye) and severely constricted visual fields. Case 2 was a 36-year-old man with 4 months of worsening headaches, reduced visual acuity (count fingers at 1 foot in each eye), severely constricted visual fields, and papilledema. Case 3 was a 39-year-old man with papilledema causing progressive vision loss (20/80 in both eyes), headaches, and relapsing limb sensorimotor deficits. Case 4 was a 19-year-old man with 3 months of progressive bilateral visual decline (20/400 in the right eye, 20/600 in the left eye), central scotoma, and optic disc pallor consistent with optic neuropathy without papilledema. All 4 patients met clinical and electrodiagnostic criteria of CIDP. Cases 3 and 4 each tested positive for serum neurofascin-155 IgG4 antibodies. All patients were managed with immunomodulatory therapy. Cases 1 and 2 also each required surgical intervention with bilateral optic nerve sheath fenestration and cerebrospinal fluid (CSF) shunting procedures. Conclusion: Vision loss from optic neuropathy with or without papilledema has rarely been reported in CIDP, and typically has been described in the context of longstanding disease. Our cases highlight how CIDP can present with early vision loss that may be profound and challenging to manage if diagnosis is delayed. CIDP should be considered in any patient with new progressive vision loss when associated with peripheral sensorimotor symptoms and elevated CSF protein. The small subgroup of CIDP patients with neurofascin-155 antibodies may be at particular risk of optic nerve involvement.

An abstract is unavailable. This article is available as a PDF only.

An abstract is unavailable. This article is available as a PDF only.