A Prediction Model to Discriminate Small Choroidal Melanoma from Choroidal Nevus

Pseudomelanoma diagnosis in a tertiary ophthalmologic centre in Spain

OBJECTIVE

To determine the types and frequency of lesions that can be misdiagnosed as choroidal or ciliary body melanomas (posterior uveal tract melanoma [PUM]).

METHODS

This is a retrospective, descriptive study examining data from patients referred to Hospital La Paz in Madrid with the diagnosis of possible PUM between January 2005 and March 2020. All patients referred for PUM were studied. In collaboration with an oncology-specialized ophthalmologic centre, each patient underwent a full ophthalmic examination, ultrasonography, and optical coherence tomography, with agreed clinical criteria used to differentiate melanomas from pseudomelanomas.

RESULTS

In our cohort of 715 patients, 48.9% had pseudomelanomas. Thirty-five different conditions were misdiagnosed as melanomas. The 5 most common conditions were choroidal nevus (40.5%), peripheral exudative hemorrhagic chorioretinopathy (12%), choroidal hemangioma (10.5%), choroidal metastasis (8%), and age-related macular degeneration (4%).

CONCLUSIONS

Altering the diagnosis and changing the treatment and prognosis for patients can be difficult for a referral centre. Herein we present the largest European cohort investigated and highlight the importance of identifying the correct diagnosis to prevent mistreatment and possible overtreatment. These misdiagnoses can have an emotional effect on patients and their families, which could be avoided with a correct diagnosis. We analyze the most common pseudomelanoma diagnoses to help physicians better diagnose patients in their care.

Artificial intelligence for ocular oncology

PURPOSE OF REVIEW

The aim of this article is to provide an update on the latest applications of deep learning (DL) and classical machine learning (ML) techniques to the detection and prognostication of intraocular and ocular surface malignancies.

RECENT FINDINGS

Most recent studies focused on using DL and classical ML techniques for prognostication purposes in patients with uveal melanoma (UM).

SUMMARY

DL has emerged as the leading ML technique for prognostication in ocular oncological conditions, particularly in UM. However, the application of DL may be limited by the relatively rarity of these conditions.

Artificial intelligence in retinal image analysis: Development, advances, and challenges

Modern advances in diagnostic technologies offer the potential for unprecedented insight into ophthalmic conditions relating to the retina. We discuss the current landscape of artificial intelligence in retina with respect to screening, diagnosis, and monitoring of retinal pathologies such as diabetic retinopathy, diabetic macular edema, central serous chorioretinopathy, and age-related macular degeneration. We review the methods used in these models and evaluate their performance in both research and clinical contexts and discuss potential future directions for investigation, use of multiple imaging modalities in artificial intelligence algorithms, and challenges in the application of artificial intelligence in retinal pathologies.

Advances in multimodal imaging for diagnosis of pigmented ocular fundus lesions

Pigmented ocular fundus lesions can range from benign to malignant. While observation is reasonable for asymptomatic benign lesions, early recognition of tumours that are vision or life threatening is critical for long-term prognosis. With recent advances and increased accessibility of multimodal imaging, it is important that providers understand how to best use these tools to detect tumours that require early referral to subspecialty centres. This review aims to provide an overview of pigmented ocular fundus lesions and their defining characteristics using multimodal imaging. We cover the spectrum of pigmented ocular fundus lesions, including freckle and focal aggregates of normal or near-normal uveal melanocytes, retinal pigment epithelium (RPE) hyperplasia, congenital hypertrophy of the RPE, RPE hamartoma associated with familial adenomatous polyposis, congenital simple hamartoma of the RPE, combined hamartoma of the retina and RPE (congenital hypertrophy of the RPE), choroidal nevus, melanocytosis, melanocytoma, melanoma, adenoma, and RPE adenocarcinoma. We describe key diagnostic features using multimodal imaging modalities of ultra-widefield fundus photography, fundus autofluorescence, optical coherence tomography (OCT), enhanced-depth imaging OCT, ultrasonography, fluorescein angiography, indocyanine green angiography, and OCT angiography (OCTA), with particular attention to diagnostic features that could be missed on fundus examination alone. Finally, we review what is on the horizon, including applications of artificial intelligence. Through skilled application of current and emerging imaging technologies, earlier detection of sight- and life-threatening melanocytic ocular fundus tumours can lead to improved patient prognosis.

Small Choroidal Melanoma: Outcomes After Surveillance Versus Immediate Treatment

PURPOSE

To quantify potential loss (loss of vision) and gain (freedom from metastasis) in patients with small choroidal melanoma treated after a period of surveillance to document growth.

METHODS

A total of 167 patients with small choroidal melanoma (size: 5.0-16.0 mm in largest basal diameter and 1.0-2.5 mm in height) were identified: 42 treated after surveillance (documented growth) and 125 treated immediately. A prediction model was applied to each patient in the immediate treatment group to obtain the predicted risk of melanoma (high risk vs low risk). Potential loss (loss of vision) and gain (freedom from metastasis) were compared between the low-risk immediate treatment group and those treated after surveillance.

RESULTS

By using the optimal cut point (0.60; 95% confidence interval: 0.37-0.61) of predicted risk for small choroidal melanoma (sensitivity: 0.74, specificity: 0.95), we identified 94 (75%) patients as high risk (score: ≥0.6) and the remaining 31 (25%) as having low-risk melanoma (score: <0.6). Over a median follow-up of 34.6 months, 5 developed metastasis (high risk = 4, low risk = 1) compared with 1 patient in the surveillance group. Initial visual acuity and loss of <15-letter visual acuity were not significantly different at 36 months between the low-risk patients immediately treated and those treated after surveillance (81% vs 83%), respectively.

CONCLUSIONS

Low-risk choroidal melanoma identified by the prediction model can be labeled as an indeterminate melanocytic tumor. Such patients can be managed by surveillance to document growth before receiving vision-threatening treatment without increased risk of metastatic death. NOTE: Publication of this article is sponsored by the American Ophthalmological Society.