An Atypical Optic Nerve Head Mass.

A 47-year-old man presented with a sudden decrease in vision in the right eye and a history of binocular diplopia, bilateral cranial nerve 6 palsies, and an undifferentiated pontine mass. Examination revealed a large optic nerve head mass with optic disc hyperemia, scattered dot hemorrhages, a placoid lesion in the posterior pole, and a mass protruding out of the optic nerve head. What would you do next?

PERFORMANCE ASSESSMENT OF AN ARTIFICIAL INTELLIGENCE CHATBOT IN CLINICAL VITREORETINAL SCENARIOS.

PURPOSE: To determine how often ChatGPT is able to provide accurate and comprehensive information regarding clinical vitreoretinal scenarios. To assess the types of sources ChatGPT primarily uses and to determine whether they are hallucinated. METHODS: This was a retrospective cross-sectional study. The authors designed 40 open-ended clinical scenarios across four main topics in vitreoretinal disease. Responses were graded on correctness and comprehensiveness by three blinded retina specialists. The primary outcome was the number of clinical scenarios that ChatGPT answered correctly and comprehensively. Secondary outcomes included theoretical harm to patients, the distribution of the type of references used by the chatbot, and the frequency of hallucinated references. RESULTS: In June 2023, ChatGPT answered 83% of clinical scenarios (33/40) correctly but provided a comprehensive answer in only 52.5% of cases (21/40). Subgroup analysis demonstrated an average correct score of 86.7% in neovascular age-related macular degeneration, 100% in diabetic retinopathy, 76.7% in retinal vascular disease, and 70% in the surgical domain. There were six incorrect responses with one case (16.7%) of no harm, three cases (50%) of possible harm, and two cases (33.3%) of definitive harm. CONCLUSION: ChatGPT correctly answered more than 80% of complex open-ended vitreoretinal clinical scenarios, with a reduced capability to provide a comprehensive response.

Bilateral peripheral ulcerative keratitis in a patient with known Sézary syndrome.

Purpose: To report a case of bilateral peripheral ulcerative keratitis (PUK) in a patient with underlying Sézary syndrome. Observations: A 58-year-old male presented with bilateral corneal ulceration with stromal thinning and was diagnosed with PUK. He was actively being treated for Sézary syndrome, a cutaneous T-cell lymphoma. He had no lagophthalmos or other adnexal abnormalities that would lead to ocular surface breakdown. A systemic autoimmune and infectious workup for PUK was unremarkable. His keratitis resolved after treatment with oral prednisone. Conclusions and importance: We describe a previously undocumented association of PUK with Sézary syndrome in a patient without adnexal disease.

Near-Surface Dose Correlates With Moist Desquamation and Unplanned Reconstructive Surgery in Patients With Implant-Based Reconstruction Receiving Postmastectomy Radiation Therapy.

Purpose: Postmastectomy radiation therapy (PMRT) reduces disease recurrence in appropriately selected patients but may compromise implant-based reconstruction. We investigated whether near-surface dose correlates with radiation-related toxic effects in these patients. Methods and Materials: Patients receiving PMRT at a single institution from 2016 to 2019 were retrospectively reviewed. Patient demographics and treatment information were collected. Three near-surface structures were retrospectively generated, bound by the chest wall tangent beam as well as the skin surface and the skin-3 mm contour (SR3), skin surface and skin-5 mm contour (SR5), or skin-5 and skin-10 mm contours. Dosimetric analysis of these near-surface contours was performed in 2 Gy intervals. Univariate and multivariate analyses were used to identify predictors of moist desquamation, grade 2+ chest wall pain, use of opiate pain medication, unplanned reconstructive surgery, and implant failure. Logistic regression for each outcome and near-surface contour was performed for receiver-operator area under the curve (AUC) analysis and the Youden J Statistic was used to determine the optimal threshold for each dosimetric parameter. Results: Of 126 patients reviewed, 109 met the study's eligibility criteria. Median follow-up was 2.3 years. Twenty-five patients (23%) underwent unplanned reconstructive surgery, and 10 (9.2%) experienced implant failure. Among clinical variables, low body mass index and history of smoking predicted unplanned surgery on univariate and multivariate analyses, and moist desquamation predicted grade 2+ chest wall pain. The top dosimetric parameters by AUC for moist desquamation, grade 2+ chest wall pain, use of opiates, unplanned reconstructive surgery, and implant failure were SR5 D10 cc (AUC = 0.701, optimal threshold 57.8 Gy, P < .001), SR3 D10 cc (AUC = 0.600, optimal threshold 56.8 Gy, P = .079), SR5 D10 cc (AUC = 0.642, optimal threshold 57.3 Gy, P = .041), SR3 V44 Gy (AUC = 0.711, optimal threshold 81%, P = .001), and SR3 V44 Gy (AUC = 0.688, optimal threshold 82%, P = .052), respectively. Conclusions: Near-surface dose correlates with moist desquamation and unplanned reconstructive surgery after PMRT. Further evaluation of prospective optimization of dosimetric parameters related to SR3 and SR5 should be considered.

LONG-TERM VISUAL OUTCOMES AND THE TIMING OF SURGICAL REPAIR OF FOVEA-SPLITTING RHEGMATOGENOUS RETINAL DETACHMENTS.

PURPOSE: To evaluate the visual outcomes and the affect of timing of surgical repair of fovea-splitting rhegmatogenous retinal detachments. METHOD: A retrospective, consecutive cohort from multiple surgeons at a single center. Fovea status (fovea-on, fovea-splitting, or fovea-off) was classified by preoperative optical coherence tomography. The primary outcome measure was the visual acuity at the last follow-up that was further correlated with the timing of surgical repair. RESULTS: One hundred and ninety-five eyes were included with 62 fovea-on, 65 fovea-splitting, and 68 fovea-off detachments. The mean preoperative logarithm of the minimum angle of resolution visual acuity for fovea-on, fovea-splitting, and fovea-off groups was 0.16 ± 0.21, 0.70 ± 0.56, and 1.67 ± 0.87, respectively (P = <0.001). Mean postoperative logarithm of the minimum angle of resolution visual acuity for fovea-on, fovea-splitting, and fovea-off groups were 0.07 ± 0.13, 0.10 ± 0.15, and 0.20 ± 0.22, respectively (P = <0.001). A statistically significant difference in mean postoperative logMAR visual acuity was found between fovea-off and fovea-on groups (P = 0.003) and between fovea-off and fovea-splitting groups (P = 0.013), however not between fovea-on and fovea-splitting groups (P = 0.827). Visual acuity improved when repair was performed earlier after presentation for fovea-on (R = 0.378, P = 0.002) and fovea-off groups (R = 0.277, P = 0.022), but not for the fovea-splitting group (R = 0.089, P = 0.481). CONCLUSION: We described the favorable visual outcomes of surgery for fovea-splitting rhegmatogenous retinal detachment and correlated these with the timing of surgical repair, which may help guide the management of this urgent, vision-threatening condition.

Mapping of Metastatic Level I Axillary Lymph Nodes in Patients with Newly Diagnosed Breast Cancer.

PURPOSE: We examined the distribution of pretreatment nodal metastases to the level I axilla (Ax-L1) to assess the appropriateness of current breast atlases and provide guidelines in relationship to easily identifiable anatomic landmarks for accurate delineation of this lymph node (LN) basin. METHODS AND MATERIALS: Patients with newly diagnosed breast cancer and biopsy-proven metastatic Ax-L1 LNs were identified. We related the location of each LN to its most adjacent rib and its distance from the bottom of the humeral head, axillary vessels, and a line connecting the anterior aspects of the pectoralis and latissimus dorsi muscles (P-L line). LNs were mapped onto a representative planning computed tomography scan, and their distribution was used to validate the current Radiation Therapy Oncology Group, European Society for Radiotherapy and Oncology, and Radiotherapy Comparative Effectiveness breast atlases. Furthermore, we examined metastases to a subregion encompassing the superolateral Ax-L1, irradiation of which correlates highly with lymphedema. RESULTS: We identified 106 eligible patients with 107 biopsied LNs. All LNs fell between the second and fifth ribs (mean, 3.8 ± 0.56). Mean distance from the inferior aspect of the humeral head was 4.3 ± 1.6 cm (range, 0.3-8.4). Mean distance from the inferior aspect of the axillary vessels was 2.9 ± 1.5 cm (range, -0.6 to 5.4). Mean distance from the P-L line was 0.01 ± 1.9 cm (range, -2.2 to 2.4); negative and positive values denote medial or lateral to the P-L line. A Radiation Therapy Oncology Group-compliant Ax-L1 consensus contour, created from contours by 4 attending breast radiation oncologists, partially or fully missed 45% of mapped LNs. European Society for Radiotherapy and Oncology- and Radiotherapy Comparative Effectiveness-compliant Ax-L1 similarly missed 46% and 34% of mapped LNs, respectively. LNs were most frequently missed in the lateral direction. The superolateral Ax-L1 encompassed 9.3% of the mapped LNs. CONCLUSIONS: A significant percentage of at-risk Ax-L1 tissue falls outside current contouring atlases. We propose expansion of the recommended Ax-L1 borders, most notably in the lateral direction.