Colour fusion effect on deep learning classification of uveal melanoma.

BACKGROUND: Reliable differentiation of uveal melanoma and choroidal nevi is crucial to guide appropriate treatment, preventing unnecessary procedures for benign lesions and ensuring timely treatment for potentially malignant cases. The purpose of this study is to validate deep learning classification of uveal melanoma and choroidal nevi, and to evaluate the effect of colour fusion options on the classification performance. METHODS: A total of 798 ultra-widefield retinal images of 438 patients were included in this retrospective study, comprising 157 patients diagnosed with UM and 281 patients diagnosed with choroidal naevus. Colour fusion options, including early fusion, intermediate fusion and late fusion, were tested for deep learning image classification with a convolutional neural network (CNN). F1-score, accuracy and the area under the curve (AUC) of a receiver operating characteristic (ROC) were used to evaluate the classification performance. RESULTS: Colour fusion options were observed to affect the deep learning performance significantly. For single-colour learning, the red colour image was observed to have superior performance compared to green and blue channels. For multi-colour learning, the intermediate fusion is better than early and late fusion options. CONCLUSION: Deep learning is a promising approach for automated classification of uveal melanoma and choroidal nevi. Colour fusion options can significantly affect the classification performance.

The 2018 World Health Organization Classification of Cutaneous, Mucosal, and Uveal Melanoma: Detailed Analysis of 9 Distinct Subtypes Defined by Their Evolutionary Pathway.

CONTEXT.­: There have been major advances in the understanding of melanoma since the last revision of the World Health Organization (WHO) classification in 2006. OBJECTIVE.­: To discuss development of the 9 distinct types of melanoma and distinguishing them by their epidemiology, clinical and histologic morphology, and genomic characteristics. Each melanoma subtype is placed at the end of an evolutionary pathway that is rooted in its respective precursor, wherever appropriate and feasible, based on currently known data. Each precursor has a variable risk of progression culminating in its fully evolved, invasive melanoma. DATA SOURCES.­: This review is based on the "Melanocytic Tumours" section of the 4th edition of the WHO Classification of Skin Tumours, published in 2018. CONCLUSIONS.­: Melanomas were divided into those etiologically related to sun exposure and those that are not, as determined by their mutational signatures, anatomic site, and epidemiology. Melanomas on the sun-exposed skin were further divided by the histopathologic degree of cumulative solar damage (CSD) of the surrounding skin, into low and high CSD, on the basis of degree of associated solar elastosis. Low-CSD melanomas include superficial spreading melanomas and high-CSD melanomas incorporate lentigo maligna and desmoplastic melanomas. The "nonsolar" category includes acral melanomas, some melanomas in congenital nevi, melanomas in blue nevi, Spitz melanomas, mucosal melanomas, and uveal melanomas. The general term melanocytoma is proposed to encompass "intermediate" tumors that have an increased (though still low) probability of disease progression to melanoma.

Prognostication of uveal melanoma is simple and highly predictive using The Cancer Genome Atlas (TCGA) classification: A review.

Purpose: The cancer genome atlas (TCGA) is a comprehensive project supported by the National Cancer Institute (NCI) in the United States to explore molecular alterations in cancer, including uveal melanoma (UM). This led to TCGA classification for UM. In this report, we review the American Joint Committee on Cancer (AJCC) classification and TCGA classification for UM from the NCI's Center for Cancer Genomics (NCI CCG) (based on enucleation specimens [n = 80 eyes]) and from Wills Eye Hospital (WEH) (based on fine needle aspiration biopsy [FNAB] specimens [n = 658 eyes]). We then compare accuracy and predictability of AJCC versus (vs.) TCGA. Methods: Review of published reports on AJCC and TCGA classification for UM was performed. Outcomes based on AJCC 7th and 8th editions were assessed. For TCGA, UM was classified based on chromosomes 3 and 8 findings including disomy 3 (D3), monosomy 3 (M3), disomy 8 (D8), 8q gain (8qG), or 8q gain multiple (8qGm) and combined into four classes including Class A (D3/D8), Class B (D3/8qG), Class C (M3/8qG), and Class D (M3/8qGm). Outcomes of metastasis and death were explored and a comparison (AJCC vs. TCGA) was performed. Results: In the NCI CCG study, there were 80 eyes with UM sampled by enucleation (n = 77), resection (n = 2), or orbitotomy (n = 1) and analysis revealed four distinct genetic classes. Metastasis and death outcomes were subsequently evaluated per class in the WEH study. The WEH study reviewed 658 eyes with UM, sampled by FNAB, and found Class A (n = 342, 52%), B (n = 91, 14%), C (n = 118, 18%), and D (n = 107, 16%). Comparison by increasing class (A vs. B vs. C vs. D) revealed older mean patient age (P < 0.001), worse entering visual acuity (P < 0.001), greater distance from the optic disc (P < 0.001), larger tumor diameter (P < 0.001), and greater tumor thickness (P < 0.001). Regarding outcomes, more advanced TCGA class demonstrated increased 5-year risk for metastasis (4% vs. 20% vs. 33% vs. 63%,P < 0.001) with corresponding increasing hazard ratio (HR) (1.0 vs. 4.1, 10.1, 30.0,P= 0.01 for B vs. A andP < 0.001 for C vs. A and D vs. A) as well as increased 5-year estimated risk for death (1% vs. 0% vs. 9% vs. 23%,P < 0.001) with corresponding increasing HR (1 vs. NA vs. 3.1 vs. 13.7,P= 0.11 for C vs. A andP < 0.001 for D vs. A). Comparison of AJCC to TCGA classification revealed TCGA was superior in prediction of metastasis and death from UM. Conclusion: TCGA classification for UM is simple, accurate, and highly predictive of melanoma-related metastasis and death, more so than the AJCC classification.

Genetic Analysis of Uveal Melanoma in 658 Patients Using the Cancer Genome Atlas Classification of Uveal Melanoma as A, B, C, and D.

PURPOSE: The Cancer Genome Atlas (TCGA) classification has been validated for uveal melanoma (UM) prognostication. We applied TCGA classification to UM biopsied using fine-needle aspiration biopsy (FNAB) to determine the predictability for metastasis and death. DESIGN: Retrospective cohort study. PARTICIPANTS: Patients with UM treated with plaque radiotherapy at Wills Eye Hospital, Philadelphia, Pennsylvania, from October 1, 2008, through December 31, 2018, who completed genetic analysis of chromosomes 3 and 8 after FNAB. METHODS: Tumors were classified as A, B, C, or D and were compared using the chi-square test, Fisher exact test, analysis of variance, and Kaplan-Meier analysis. MAIN OUTCOME MEASURES: Metastasis and death. RESULTS: Six hundred fifty-eight UM patients were categorized accordingly as TCGA class A (n = 342 [52%]), B (n = 91 [14%]), C (n = 118 [18%]), and D (n = 107 [16%]). More advanced tumor classification revealed older mean patient age (56 vs. 53 vs. 60 vs. 63 years, respectively; P < 0.001), worse presenting visual acuity (20/20-20/50: 81% vs. 67% vs. 71% vs. 66%, respectively; P < 0.001), greater distance from the optic disc (3.5 vs. 4.9 vs. 5.7 vs. 5.3 mm, respectively; P < 0.001), larger tumor basal diameter (10.3 vs. 12.9 vs. 13.9 vs. 15.3 mm, respectively; P < 0.001), and greater tumor thickness (4.3 vs. 6.1 vs. 6.6 vs. 7.5 mm, respectively; P < 0.001). After mean follow-up (47.6 vs. 47.6 vs. 42.9 vs. 28.7 months, respectively; P < 0.001), more advanced TCGA class was associated with increased risk of metastasis (3% vs. 10% vs. 25% vs. 41%, respectively; P < 0.001) and death (1% vs. 0% vs. 3% vs. 9%, respectively; P < 0.001). Compared with class A, the 5-year hazard ratio for metastasis increased at 4.1 (B vs. A; P = 0.01), 10.1 (C vs. A; P < 0.001), and 30.0 (D vs. A; P < 0.001). The 5-year hazard ratio for death increased at 3.1 (C vs. A; P = 0.11) and 13.7 (D vs. A; P < 0.001) with no deaths in class B. CONCLUSIONS: Grouping of UM using TCGA classification predicts the risk of melanoma-related metastasis and death.

Modifications of Callender's Classification of Uveal Melanoma at the Armed Forces Institute of Pathology.

One hundred well-documented cases of uveal melanoma accessioned at the Armed Forces Institute of Pathology before 1970 were reviewed and reclassified to identify changes made in the Callender classification. We compared the new classification with the original classification to determine the effect of the changes on the prediction of outcome for the patient after enucleation. Staff pathologists had originally classified 52 of the 100 cases as spindle-cell type melanoma. Only 31 of the 100 cases were reclassified as spindle-cell types (two spindle-cell nevi and 29 spindle-cell melanomas). Tumors classified as mixed-cell type were further subdivided into groups based on the percentage and size of the epithelioid cells. Tumors formerly classified as spindle-cell type that contained small or rare epithelioid cells were reclassified as mixed-cell type. This improved the prediction of outcome for the patient. We found that nucleolar size and pleomorphism are important variables that should be considered in the classification of uveal melanomas.

Integrated Genomic Classification of Melanocytic Tumors of the Central Nervous System Using Mutation Analysis, Copy Number Alterations, and DNA Methylation Profiling.

Purpose: In the central nervous system, distinguishing primary leptomeningeal melanocytic tumors from melanoma metastases and predicting their biological behavior solely using histopathologic criteria may be challenging. We aimed to assess the diagnostic and prognostic value of integrated molecular analysis.Experimental Design: Targeted next-generation sequencing, array-based genome-wide methylation analysis, and BAP1 IHC were performed on the largest cohort of central nervous system melanocytic tumors analyzed to date, including 47 primary tumors of the central nervous system, 16 uveal melanomas, 13 cutaneous melanoma metastases, and 2 blue nevus-like melanomas. Gene mutation, DNA-methylation, and copy-number profiles were correlated with clinicopathologic features.Results: Combining mutation, copy-number, and DNA-methylation profiles clearly distinguished cutaneous melanoma metastases from other melanocytic tumors. Primary leptomeningeal melanocytic tumors, uveal melanomas, and blue nevus-like melanoma showed common DNA-methylation, copy-number alteration, and gene mutation signatures. Notably, tumors demonstrating chromosome 3 monosomy and BAP1 alterations formed a homogeneous subset within this group.Conclusions: Integrated molecular profiling aids in distinguishing primary from metastatic melanocytic tumors of the central nervous system. Primary leptomeningeal melanocytic tumors, uveal melanoma, and blue nevus-like melanoma share molecular similarity with chromosome 3 and BAP1 alterations, markers of poor prognosis. Clin Cancer Res; 24(18); 4494-504. ©2018 AACR.

polyClustR: defining communities of reconciled cancer subtypes with biological and prognostic significance.

BACKGROUND: To ensure cancer patients are stratified towards treatments that are optimally beneficial, it is a priority to define robust molecular subtypes using clustering methods applied to high-dimensional biological data. If each of these methods produces different numbers of clusters for the same data, it is difficult to achieve an optimal solution. Here, we introduce "polyClustR", a tool that reconciles clusters identified by different methods into subtype "communities" using a hypergeometric test or a measure of relative proportion of common samples. RESULTS: The polyClustR pipeline was initially tested using a breast cancer dataset to demonstrate how results are compatible with and add to the understanding of this well-characterised cancer. Two uveal melanoma datasets were then utilised to identify and validate novel subtype communities with significant metastasis-free prognostic differences and associations with known chromosomal aberrations. CONCLUSION: We demonstrate the value of the polyClustR approach of applying multiple consensus clustering algorithms and systematically reconciling the results in identifying novel subtype communities of two cancer types, which nevertheless are compatible with established understanding of these diseases. An R implementation of the pipeline is available at: https://github.com/syspremed/polyClustR.

Association between traditional clinical high-risk features and gene expression profile classification in uveal melanoma.

PURPOSE: To evaluate the association between traditional clinical high-risk features of uveal melanoma patients and gene expression profile (GEP). METHODS: This was a retrospective, single-center, case series of patients with uveal melanoma. Eighty-three patients met inclusion criteria for the study. Patients were examined for the following clinical risk factors: drusen/retinal pigment epithelium (RPE) changes, vascularity on B-scan, internal reflectivity on A-scan, subretinal fluid (SRF), orange pigment, apical tumor height/thickness, and largest basal dimensions (LBD). A novel point system was created to grade the high-risk clinical features of each tumor. Further analyses were performed to assess the degree of association between GEP and each individual risk factor, total clinical risk score, vascularity, internal reflectivity, American Joint Committee on Cancer (AJCC) tumor stage classification, apical tumor height/thickness, and LBD. RESULTS: Of the 83 total patients, 41 were classified as GEP class 1A, 17 as class 1B, and 25 as class 2. The presence of orange pigment, SRF, low internal reflectivity and vascularity on ultrasound, and apical tumor height/thickness ≥ 2 mm were not statistically significantly associated with GEP class. Lack of drusen/RPE changes demonstrated a trend toward statistical association with GEP class 2 compared to class 1A/1B. LBD and advancing AJCC stage was statistically associated with higher GEP class. CONCLUSIONS: In this cohort, AJCC stage classification and LBD were the only clinical features statistically associated with GEP class. Clinicians should use caution when inferring the growth potential of melanocytic lesions solely from traditional funduscopic and ultrasonographic risk factors without GEP data.

Gene expression profiling in uveal melanoma: technical reliability and correlation of molecular class with pathologic characteristics.

BACKGROUND: A 15-gene expression profile test has been clinically validated and is widely utilized in newly diagnosed uveal melanoma (UM) patients to assess metastatic potential of the tumor. As most patients are treated with eye-sparing radiotherapy, there is limited tumor tissue available for testing, and technical reliability and success of prognostic testing are critical. This study assessed the analytical performance of the 15-gene expression test for UM and the correlation of molecular class with pathologic characteristics. METHODS: Inter-assay, intra-assay, inter-instrument/operator, and inter-site experiments were conducted, and concordance of the 15-gene expression profile test results and associated discriminant scores for matched tumor samples were evaluated. Technical success was determined from de-identified clinical reports from January 2010 - May 2016. Pathologic characteristics of enucleated tumors were correlated with molecular class results. RESULTS: Inter-assay concordance on 16 samples run on 3 consecutive days was 100%, and matched discriminant scores were strongly correlated (R2 = 0.9944). Inter-assay concordance of 46 samples assayed within a one year period was 100%, with an R2 value of 0.9747 for the discriminant scores. Intra-assay concordance of 12 samples run concurrently in duplicates was 100%; discriminant score correlation yielded an R2 of 0.9934. Concordance between two sites assessing the same tumors was 100% with an R2 of 0.9818 between discriminant scores. Inter-operator/instrument concordance was 96% for Class 1/2 calls and 90% for Class 1A/1B calls, and the discriminant scores had a correlation R2 of 0.9636. Technical success was 96.3% on 5516 samples tested since 2010. Increased largest basal diameter and thickness were significantly associated with Class 1B and Class 2 vs. Class 1A signatures. CONCLUSIONS: These results show that the 15-gene expression profile test for UM has robust, reproducible performance characteristics. The technical success rate during clinical testing remains as high as first reported during validation. As molecular testing becomes more prevalent for supporting precision medicine efforts, high technical success and reliability are key characteristics when testing such limited and precious samples. The performance of the 15-gene expression profile test in this study should provide confidence to physicians who use the test's molecular classification to inform patient management decisions.

Integrative Analysis Identifies Four Molecular and Clinical Subsets in Uveal Melanoma.

Comprehensive multiplatform analysis of 80 uveal melanomas (UM) identifies four molecularly distinct, clinically relevant subtypes: two associated with poor-prognosis monosomy 3 (M3) and two with better-prognosis disomy 3 (D3). We show that BAP1 loss follows M3 occurrence and correlates with a global DNA methylation state that is distinct from D3-UM. Poor-prognosis M3-UM divide into subsets with divergent genomic aberrations, transcriptional features, and clinical outcomes. We report change-of-function SRSF2 mutations. Within D3-UM, EIF1AX- and SRSF2/SF3B1-mutant tumors have distinct somatic copy number alterations and DNA methylation profiles, providing insight into the biology of these low- versus intermediate-risk clinical mutation subtypes.

Melanoma subtypes demonstrate distinct PD-L1 expression profiles.

PD-L1 expression in the tumor immune microenvironment is recognized as both a prognostic and predictive biomarker in patients with cutaneous melanoma, a finding closely related to its adaptive (IFN-γ-mediated) mechanism of expression. Approximately 35% of cutaneous melanomas express PD-L1, however, the expression patterns, levels, and prevalence in rarer melanoma subtypes are not well described. We performed immunohistochemistry for PD-L1 and CD8 on 200 formalin-fixed paraffin-embedded specimens from patients with acral (n=16), mucosal (n=36), uveal (n=103), and chronic sun-damaged (CSD) (n=45) melanomas (24 lentigo maligna, 13 'mixed' desmoplastic, and 8 'pure' desmoplastic melanomas). CD8+ tumor-infiltrating lymphocyte (TIL) densities were characterized as mild, moderate, or severe, and their geographic association with PD-L1 expression was evaluated. Discrete lymphoid aggregates, the presence of a spindle cell morphology, and the relationship of these features with PD-L1 expression were assessed. PD-L1 expression was observed in 31% of acral melanomas, 44% of mucosal melanomas, 10% of uveal melanomas, and 62% of CSD melanomas (P<0.0001). Compared to our previously characterized cohort of cutaneous melanomas, the proportion of PD-L1(+) tumors was lower in uveal (P=0.0002) and higher in CSD (P=0.0073) melanomas, while PD-L1 expression in the acral and mucosal subtypes was on par. PD-L1 expression in all subtypes correlated with a moderate-severe grade of CD8+ TIL (all, P<0.003), supporting an adaptive mechanism of expression induced during the host antitumor response. The tumor microenvironments observed in CSD melanomas segregated by whether they were the pure desmoplastic subtype, which showed lower levels of PD-L1 expression when compared to other CSD melanomas (P=0.047). The presence of lymphoid aggregates was not associated with the level of PD-L1 expression, while PD-L1(+) cases with spindle cell morphology demonstrated higher levels of PD-L1 than those with a nested phenotype (P<0.0001). Our findings may underpin the reported clinical response rates for anti-PD-1 monotherapy, which vary by subtype.

Comparison of Alternative Tumor Size Classifications for Posterior Uveal Melanomas.

Purpose: Determine which posterior uveal melanoma (PUM) size classification with three categories has the best prognostic discrimination. Methods: Single-institution study of 424 consecutive patients with PUM. The tumor's largest basal diameter (LBD), smallest basal diameter (SBD), and thickness (TH) were estimated by fundus mapping and ultrasonography. Tumors were assigned to "small," "medium," or "large" size categories defined by 11 different classifications (Linear LBD, Rectangular LBD × TH, Cubic LBD × SBD × TH, Warren Original, Warren Modified, Augsburger, COMS Original, COMS Revised, TNM 2002, and modified TNM 2010 classification [a,b]). Prognostic significance of classifications was evaluated by Kaplan-Meier event curves with computation of log rank test for trend statistic. Results: In six classification systems (Warren Original, Warren Modified, COMS Revised, TNM 2002, TNM 2010a, TNM 2010b) >50% of tumors fell within one subgroup. In the Warren Original classification <5% of tumors fell within one subgroup. Separation of Kaplan-Meier curves among three size categories was judged "excellent" in four classifications (Linear LBD, Cubic Volume, TNM 2010a, and TNM 2010b) and "very poor" in the Warren Original. Linear LBD classification was associated with highest log rank statistic value. TNM 2010a, TNM 2010b, TNM 2002, Augsburger, and Cubic Volume classifications were also determined to be quite good. Conclusions: Linear LBD classification was the best three-size category discriminator among low-, intermediate-, and high-risk subgroups. Considering our findings, it seems possible that the arduous work required to apply complex classifications, especially for three-category systems, for PUM may not be justified in routine clinical practice.

Focus on cutaneous and uveal melanoma specificities.

Cutaneous melanoma (CM) and uveal melanoma (UM) derive from cutaneous and uveal melanocytes that share the same embryonic origin and display the same cellular function. However, the etiopathogenesis and biological behaviors of these melanomas are very different. CM and UM display distinct landscapes of genetic alterations and show different metastatic routes and tropisms. Hence, therapeutic improvements achieved in the last few years for the treatment of CM have failed to ameliorate the clinical outcomes of patients with UM. The scope of this review is to discuss the differences in tumorigenic processes (etiologic factors and genetic alterations) and tumor biology (gene expression and signaling pathways) between CM and UM. We develop hypotheses to explain these differences, which might provide important clues for research avenues and the identification of actionable vulnerabilities suitable for the development of new therapeutic strategies for metastatic UM.

PRAME as an Independent Biomarker for Metastasis in Uveal Melanoma.

PURPOSE: Uveal melanoma (UM) can be classified by gene expression profiling (GEP) into Class 1 (low metastatic risk) and Class 2 (high metastatic risk), the latter being strongly associated with mutational inactivation of the tumor suppressor BAP1. Nevertheless, a small percentage of Class 1 tumors give rise to metastatic disease. The purpose of this study was to identify biomarkers of metastasis in Class 1 tumors. EXPERIMENTAL DESIGN: A total of 389 consecutive patients with UM were assigned to Class 1 or Class 2 using a prospectively validated 12-gene prognostic classifier. Selected tumors were further analyzed using global GEP and single nucleotide polymorphism microarrays. PRAME (preferentially expressed antigen in melanoma) mRNA expression was analyzed in 64 Class 1 tumors by qPCR. RESULTS: Among Class 1 UMs, the most significant predictor of metastasis was PRAME mRNA expression (P = 0.0006). The 5-year actuarial rate of metastasis was 0% for Class1(PRAME-), 38% for Class1(PRAME+), and 71% for Class 2 tumors. Median metastasis-free survival for Class1(PRAME+) patients was 88 months, compared to 32 months for Class 2 patients. Findings were validated using three independent datasets, including one using disomy 3 to identify low-risk UM. Chromosome copy number changes associated with Class1(PRAME+) tumors included gain of 1q, 6p, 8q, and 9q and loss of 6q and 11q. PRAME expression was associated with larger tumor diameter (P = 0.05) and SF3B1 mutations (P = 0.003). CONCLUSIONS: PRAME is an independent prognostic biomarker in UM, which identifies increased metastatic risk in patients with Class 1 or disomy 3 tumors. This finding may further enhance the accuracy of prognostic testing and precision medicine for UM.

Uveal Melanomas with SF3B1 Mutations: A Distinct Subclass Associated with Late-Onset Metastases.

PURPOSE: To investigate the prevalence and prognostic value of SF3B1 and EIF1AX mutations in uveal melanoma (UM) patients. DESIGN: Case series. PARTICIPANTS: Cohort of 151 patients diagnosed with and treated for UM. METHODS: SF3B1 and EIF1AX mutations in primary tumors were investigated using whole-exome sequencing (n = 25) and Sanger sequencing (n = 151). For the detection of BAP1 mutations, a previously reported cohort of 90 patients was extended using BAP1 sequencing or immunohistochemistry. MAIN OUTCOME MEASURES: The status of SF3B1, EIF1AX, and BAP1 in tumors of patients were correlated to clinical, histopathologic, and genetic parameters. Survival analyses were performed for patients whose tumors had SF3B1, EIF1AX, and BAP1 mutations. RESULTS: Patients with tumors harboring EIF1AX mutations rarely demonstrated metastases (2 of 28 patients) and overall had a longer disease-free survival (DFS; 190.1 vs. 100.2 months; P < 0.001). Within the patient group with disomy 3, UM patients with an SF3B1 mutation had an increased metastatic risk compared with those without an SF3B1 mutation (DFS, 132.8 vs. 174.4 months; P = 0.008). Patients with such a mutation were more prone to demonstrate late metastases (median, 8.2 years; range, 23-145 months). Patients with UM and loss of BAP1 expression had a significantly decreased survival (DFS, 69.0 vs. 147.9 months; P < 0.001). CONCLUSIONS: According to our data, patients with UM can be classified into 3 groups, of which EIF1AX-mutated tumors and tumors without BAP1, SF3B1, or EIF1AX mutations are associated with prolonged survival and low metastatic risk, SF3B1-mutated tumors are associated with late metastasis, and tumors with an aberrant BAP1 are associated with an early metastatic risk and rapid decline in patient DFS.

Independent Prognostic Significance of Gene Expression Profile Class and Largest Basal Diameter of Posterior Uveal Melanomas.

PURPOSE: To determine whether any conventional clinical prognostic factors for metastasis from uveal melanoma retain prognostic significance in multivariate models incorporating gene expression profile (GEP) class of the tumor cells. DESIGN: Prospective, interventional case series with a prognostic model. METHODS: Single-institution study of GEP testing and other conventional prognostic factors for metastasis and metastatic death in 299 patients with posterior uveal melanoma evaluated by fine-needle aspiration biopsy (FNAB) at the time of or shortly prior to initial treatment. Univariate prognostic significance of all evaluated potential prognostic variables (patient age, largest linear basal diameter of tumor [LBD], tumor thickness, intraocular location of tumor, melanoma cytomorphologic subtype, and GEP class) was performed by comparison of Kaplan-Meier event rate curves and univariate Cox proportional hazards modeling. Multivariate prognostic significance of combinations of significant prognostic factors identified by univariate analysis was performed using step-up and step-down Cox proportional hazards modeling. RESULTS: GEP class was the strongest prognostic factor for metastatic death in this series. However, tumor LBD, tumor thickness, and intraocular tumor location also proved to be significant individual prognostic factors in this study. On multivariate analysis, a 2-term model that incorporated GEP class and largest basal diameter was associated with strong independent significance of each of the factors. CONCLUSION: Although GEP test is the most robust prognostic indicator in uveal melanoma and early studies of mostly larger tumors found that no clinicopathologic factors had significant prognostic value independent of GEP, our single-center study, which included a substantial proportion of smaller tumors, showed that both GEP and LBD of the tumor are independent prognostic factors for metastasis and metastatic death in multivariate analysis.

American Joint Committee on Cancer Classification of Uveal Melanoma (Anatomic Stage) Predicts Prognosis in 7,731 Patients: The 2013 Zimmerman Lecture.

PURPOSE: To analyze the clinical features and prognosis of posterior uveal melanoma based on the American Joint Committee on Cancer (AJCC) (7th edition) tumor staging. DESIGN: Retrospective interventional case series. PARTICIPANTS: A total of 7731 patients. METHODS: Uveal melanoma management. MAIN OUTCOME MEASURES: Melanoma-related metastasis and death. RESULTS: Of 7731 patients with posterior uveal (ciliary body and choroidal) melanoma, the AJCC tumor staging was stage I in 2767 (36%), stage II in 3735 (48%), stage III in 1220 (16%), and stage IV in 9 (<1%). Based on tumor staging (I, II, III, and IV), features that showed significant increase with tumor staging included age at presentation (57, 58, 60, 60 years) (P < 0.001), tumor base (8, 12, 17, 17 mm) (P < 0.001), tumor thickness (2.9, 6.0, 10.1, 10.2 mm) (P < 0.001), distance to optic disc (3, 5, 5, 5 mm) (P < 0.001), distance to foveola (3, 5, 5, 5 mm) (P < 0.001), mushroom configuration (6%, 24%, 34%, 33%) (P < 0.001), plateau configuration (3%, 4%, 7%, 11%) (P < 0.001), tumor pigmentation (48%, 53%, 69%, 78%) (P < 0.001), and extraocular extension (0%, 1%, 11%, 22%) (P < 0.001). After therapy, Kaplan-Meier estimates of metastasis at 1, 5, 10, and 20 years were <1%, 5%, 12%, and 20% for stage I, 2%; 17%, 29%, and 44% for stage II; 6%, 44%, 61%, and 73% for stage III, and 100% by 1 year for stage IV. Kaplan-Meier estimates of death at 1, 5, 10, and 20 years were <1%, 3%, 6%, and 8% for stage I; <1%, 9%, 15%, and 24% for stage II; 3%, 27%, 39%, and 53% for stage III, and 100% by 1 year for stage IV. Compared with stage I, the hazard ratio for metastasis/death was 3.1/3.1 for stage II and 9.3/10.1 for stage III. CONCLUSIONS: Compared with uveal melanoma classified as AJCC stage I, the rate of metastasis/death was 3 times greater for stage II, 9 to 10 times greater for stage III, and further greater for stage IV. Early detection of posterior uveal melanoma, at a point when the tumor is small, can be lifesaving.

International Validation of the American Joint Committee on Cancer's 7th Edition Classification of Uveal Melanoma.

IMPORTANCE: Although an accurate uveal melanoma staging system is needed to improve research and patient care, the evaluation of eye cancer staging systems requires international multicenter data sharing to acquire a statistically significant analysis. OBJECTIVE: To assess patient mortality outcomes associated with uveal melanoma staging according to the 7th edition of the American Joint Committee on Cancer's AJCC Cancer Staging Manual. DESIGN, SETTING, PARTICIPANTS: A committee was formed to create patient-specific data fields for patients with uveal melanoma. Ten subspecialty ophthalmic oncology centers from 4 continents shared data. Patient selection criteria included diagnosis of uveal melanoma from April 1, 2001, to April 1, 2011, adequate records to allow tumor staging by the AJCC criteria, and follow-up for metastatic melanoma. INTERVENTIONS: Primary treatments included local resection, radiation therapy, and enucleation. MAIN OUTCOMES AND MEASURES: Metastasis after initial tumor staging with 5- and 10-year Kaplan-Meier metastasis-free point estimates, depending on AJCC prognostic stages I through IV, tumor size category, and subclassification (defined by the presence of ciliary body involvement and/or extrascleral extension). RESULTS: A total of 3809 patients were entered into the database. Of these, 3377 records (88.7%) were complete. Primary ciliary body and choroidal melanoma was the diagnosis for 3217, and 160 had primary iris melanoma. Tumor size categories were T1 in 1115 (34.7%) of the 3217 patients, T2 in 1128 patients (35.1%), T3 in 789 patients (24.5%), and T4 in 185 patients (5.8%). The 5- and 10-year Kaplan-Meier metastasis-free point estimates by tumor size categories were 97% (95% CI, 95%-98%) and 94% (95% CI, 91%-96%) for T1 tumors, 85% (95% CI, 82%-88%) and 80% (95% CI, 75%-84%) for T2 tumors, 77% (95% CI, 73%-80%) and 68% (95% CI, 60%-74%) for T3 tumors, and 61% (95% CI, 49%-71%) (5-year only) for T4 tumors, respectively. Increasing tumor size was consistent with increased metastasis risk (P < .001). Subclassifications were significantly associated with increased risk of metastasis (P < .001). The AJCC prognostic and anatomical groupings were as follows: stage I, 1030 (32.0%); stage IIA, 1095 (34.0%); stage IIB, 710 (22.1%); stage IIIA, 282 (8.8%); stage IIIB, 79 (2.5%); and stage IIIC, 21 (0.7%). The 5- and 10-year Kaplan-Meier metastasis-free estimates for prognostic stages were 97% (95% CI, 95%-98%) and 94% (95% CI, 91%-96%) for stage I, 89% (95% CI, 86%-91%) and 84% (95% CI, 80%-88%) for stage IIA, 79% (95% CI, 75%-83%) and 70% (95% CI, 62%-76%) for stage IIB, 67% (95% CI, 59%-73%) and 60% (95% CI, 51%-68%) for stage IIIA, 50% (95% CI, 33%-65%) and 50% (95% CI, 33%-65%) for stage IIIB, and 25% (95% CI, 4%-53%) (5-year only) for stage IIIC, respectively. The 160 iris melanomas were too few for subgroup analysis. CONCLUSIONS AND RELEVANCE: Multicenter, worldwide, Internet-based data sharing was used to study a heterogenous patient population in ophthalmic oncology. Our results support the continued use of the 7th edition of the AJCCCancer Staging Manual for uveal melanoma.

TUMORS OF THE NONPIGMENTED EPITHELIUM OF THE CILIARY BODY: The Lorenz E. Zimmerman Tribute Lecture.

BACKGROUND: In 1970, Dr. Lorenz Zimmerman delivered the Norman McAlister Gregg Lecture entitled "The remarkable polymorphism of tumors of the ciliary epithelium." Therein, he proposed a classification of these tumors that included congenital lesions (mainly medulloepithelioma) and acquired lesions (mainly adenoma and adenocarcinoma). The classification was based on histopathologic observations without detailed clinical information. METHODS: Review of the published literature and personal experience with tumors of the nonpigmented ciliary body epithelium. RESULTS: Since 1970, further observations through clinical examination and advanced testing with ultrasound biomicroscopy, anterior segment optical coherence tomography, and magnetic resonance imaging have expanded our knowledge regarding tumors of the nonpigmented ciliary body epithelium. Regarding medulloepithelioma, we have learned of the common associated features of neovascular glaucoma, retrolenticular neoplastic or vascular cyclitic membrane, intralesional cysts, response to radiotherapy, and association with Dicer-1 mutation. Regarding adenoma/adenocarcinoma, improved management with surgical resection (sparing globe) can be achieved. Fuchs adenoma, also termed coronal adenoma, is commonly found histopathologically, despite its rare clinical visualization, and should be added to the Zimmerman classification. CONCLUSION: Since Zimmerman's report on histopathologic features of tumors of the nonpigmented ciliary body epithelium, there have been numerous publications and further observations on the clinical features and management of these intriguing neoplasms.

A molecular revolution in uveal melanoma: implications for patient care and targeted therapy.

Uveal melanoma is the most common primary intraocular malignancy and has a strong propensity for fatal metastasis. Recent advances in the molecular genetics of uveal melanoma are revolutionizing our understanding of this cancer and the care of patients. The development of a new molecular classification of uveal melanoma based on a widely available 15-gene expression profile now allows patients at high risk of metastasis to be identified early so that individualized management can be offered. The recent discovery of major driver mutations in uveal melanoma provide a rational basis for development of new targeted therapies. Taken together, these advances are transforming our understanding and management of uveal melanoma with the ultimate goal of improving patient outcomes.