A higher AJCC stage is associated with increased occurrence of metastasis (P < .001).
A, Patients classified by AJCC tumor size; B, patients classified by CBI and/or EXE category. For CBI and/or EXE category, a indicates no CBI or EXE; b, CBI; c, EXE; and d, CBI and EXE. Higher tumor size (P < .001) and CBI and/or EXE category (P < .001) are associated with an increased occurrence of metastasis.
eTable 1. Disease Characteristics and Specimen Details for 3217 Patients With Primary Ciliary Body and Choroidal Melanoma
eTable 2. Risk of Metastasis Compared With AJCC Stage I and Tumor Size Category 1, Using a Cox Regression Model for Patients With Primary Ciliary Body and Choroidal Melanoma
eTable 3. Distribution of 3217 Patients With Primary Ciliary Body and Choroidal Melanoma by the 7th edition AJCC Tumor Size Category Matrix
eTable 4. Kaplan-Meier Metastasis-Free Point Estimates According to Size of Extraocular Extension (EXE) for 3217 Patients With Primary Ciliary Body and Choroidal Melanoma
eFigure. Kaplan-Meier Curves of Metastasis-Free Point Estimates for 3217 Patients, Classified by Size of Extraocular Extension (EXE)
. International Validation of the American Joint Committee on Cancer’s 7th Edition Classification of Uveal Melanoma. JAMA Ophthalmol. 2015;133(4):376-383. doi:10.1001/jamaophthalmol.2014.5395
Copyright 2015 American Medical Association. All Rights Reserved. Applicable FARS/DFARS Restrictions Apply to Government Use.
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.
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.
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).
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.
Staging systems serve as a foundation for cancer care.1 They provide a scientific-medical language that defines the local, regional, and systemic status of a patient’s cancer at a specific time.1 The American Joint Committee on Cancer (AJCC)/International Union Against Cancer (UICC) staging system, called TNM, is defined by the primary tumor (T), lymph node extension (N), and distant metastasis (M). In practice, TNM staging defines the corporal tumor extent.2 It is used (through translation of the TNM stages) to predict a patient’s initial prognosis, guide treatment decisions, and redefine tumor extent during a patient’s life.3
Staging helps to determine diagnostic and therapeutic options. Follow-up staging data provide statistically significant answers to questions about the efficacy of diagnostic and therapeutic choices.1- 3 Staging data analysis enables cancer caregivers to select the best available methods. In ophthalmic oncology, staging has and will continue to contribute to improvements in patient care, vision retention, and survival.
In 2009, the 7th edition of the AJCC Cancer Staging Manual provided the first widely accepted staging system for eye cancer.4 This work was created by 47 eye cancer specialists from 11 countries. During the production of the manual chapters, each eye cancer committee was tasked with using the best available evidence. For uveal melanoma, aspects of the Collaborative Ocular Melanoma Study (COMS), chapters from the 6th edition of the AJCC Cancer Staging Manual, and radiation oncology and medical physics guidelines were incorporated. However, it was the European Ophthalmic Oncology Group (EOOG) study of 7369 patients with posterior uveal melanoma that became a framework component for uveal melanoma tumor size categories and stages in the 7th edition of the AJCC Cancer Staging Manual.5,6
The staging system described in the 7th edition of the AJCC Cancer Staging Manual has become universal. It has been recognized by all cancer-related participating organizations, including more than 140 worldwide cancer registries (including those within the UICC). It has been incorporated into the instructions for authors of more than 12 major ophthalmology journals and thus their associated societies.5,7- 22 In sum, the AJCC staging system has become an international language for eye cancer specialists and their patients.
However, all cancer-staging systems must be tested. To date, one study10 has validated the prognostic value of extraocular extension (EXE) of uveal melanoma, and another study18 found AJCC tumor size categories to be predictive of posterior choroidal melanoma metastasis. In contrast, we present the first worldwide, Internet-based data-sharing study, to our knowledge, that evaluates the prognosis of patients with uveal (iris, ciliary body, and choroidal) melanoma based on the tumor size classifications, anatomical subcategories, and prognostic stages (stages I-IV) in the 7th edition of the AJCC Cancer Staging Manual.
The AJCC Ophthalmic Oncology Task Force (OOTF) collaborated to develop an Internet-based, retrospective registry to evaluate the staging system for uveal melanoma in the 7th edition of the AJCC Cancer Staging Manual. Through a consensus process, uveal melanoma OOTF committee members (primarily ophthalmic oncologists and pathologists) developed epidemiologic, clinical, and pathologic data fields. However, although the registry recorded a broad spectrum of tumor data points, the scope of this study was limited to evaluation of the TNM with regard to risk of metastasis.
Secure data storage met international standards for patient privacy protection and statistical analysis. Security measures included the lack of personal patient identifiers, SSL (Secure Sockets Layer) encryption, protection against SQL (Structured Query Language) injection, variable and session management, record locking, and trail auditing (eg, failed login attempts and webpage accessing). In addition, access to the online survey required user accounts issued by the coordinating center. Each center could only access records pertinent to their site.
Each participating site obtained local ethics and institutional review board approval for participation and entry of patient data into the online database. Informed consent was not required by the institutional review boards. When documentation of the local ethics approval was received by the coordinating center, unique login passwords were provided to initiate patient entry. All data were deidentified at the local center, where study identifiers for each patient were randomly generated. This study conforms to the Declaration of Helsinki and Health Insurance Portability and Accountability Act of 1996.
The ophthalmic oncology centers that participated in the original framework component EOOG study were excluded to provide an independent validation.6 Patient eligibility criteria included patients with primary melanoma of the iris, ciliary body, or choroid consecutively diagnosed and treated in a 10-year span from April 1, 2001, to April 1, 2011. There were no limitations regarding the method of primary tumor treatment. Patients were excluded from analysis if they had stage IV disease (metastasis) at the time of initial staging or key variables were missing or inconsistent.
The diagnosis of uveal melanoma was made by clinical and/or pathologic analysis. The clinical diagnosis of the largest basal diameter, tumor thickness, ciliary body involvement (CBI), and EXE were determined by standard clinical methods.
Mortality was defined as the time to metastatic melanoma. Metastasis was used as a surrogate for mortality based on the current assumption that uveal melanoma metastasis ultimately leads to metastatic death and the recognized difficulties in confirming the actual cause of death. There were no limitations on the methods of screening for metastasis. However, these methods included abdominal ultrasonography, computed tomography, magnetic resonance imaging, or whole-body positron emission tomography–computed tomography as was the custom and practice at each center.
The tumor size, node, and metastasis, as well as the anatomical and prognostic groups, were defined in accordance with and can be found within the 7th edition of the AJCC Cancer Staging Manual on uveal melanoma (Table 1).7 Note that in this manual there are no prognostic stages for iris melanoma.4
Continuous variables were described using medians, ranges, and interquartile ranges (IQRs), and categorical variables were described using frequencies and proportions. Time to metastasis was measured from the date of staging until the date of metastasis; patients who were alive and metastasis free at last follow-up were censored. Twenty-five patients (with primary ciliary body and/or choroidal melanoma) died without metastasis. Because the number of patients in this latter category is small relative to the size of the data set, these patients were censored at their date of death rather than treating their death as a competing risk event.23 Log-rank tests for trend, Kaplan-Meier plots, and Cox proportional hazards models were implemented to test for evidence suggesting that stage, tumor size category, or subclassification based on CBI or EXE are related to metastasis. Metastasis-free point estimates at 1, 5, and 10 years were tabulated, and in instances in which at least 10 patients remained at risk, the values are labeled in each table.
The survival package in R software, version 2.15.1 (R Foundation for Statistical Computing), was used to generate Kaplan-Meier plots, and SAS statistical software, version 9.3, TS Level 1M1 (SAS Institute Inc) was used to perform all other statistical analyses. Statistical significance was set at .05, and no adjustments were made for multiple tests.
Ten ophthalmic oncology centers from 8 countries (Argentina, Canada, Japan, Russia, Spain, Sweden, the Netherlands, and the United States) in 4 continents (North and South America, Europe, and Asia) participated in this study. A total of 3809 patients with uveal melanoma were entered into the online database. There were 181 patient entries eliminated because of lack of follow-up. Of these, 67 were eliminated because of initial stage IV disease (metastasis) and 184 because of missing or inconsistent key variables. Of these eliminated patients, 118 had unknown metastasis status, 35 had unknown EXE status, 11 had unknown CBI status, and 128 had incomplete or inconsistent data that prevented being assigned a tumor size category or prognostic stage. Some patient entries were eliminated because of more than one of these reasons. A resultant total of 3377 patient entries (88.7%) were analyzed. Primary ciliary body and choroidal melanoma was the diagnosis for 3217 patients, and 160 had primary iris melanoma.
Median follow-up (date of staging to date of last visit) was 3.2 years (IQR, 1.7-5.8 years; range, 0.08-12.6 years). Median tumor thickness was 4.7 mm (IQR, 3.0-7.1 mm; range, 1.1-23 mm), and median largest basal diameter was 11.8 mm (IQR, 9-14 mm; range, 2-30 mm). Patient and tumor demographics are listed in eTable 1 in the Supplement.
According to tumor size, 1115 (34.7%) of the 3217 patients were classified as category T1, 1128 (35.1%) as T2, 789 (24.5%) as T3, and 185 (5.8%) as T4. Tumors were further classified by subcategories (CBI and EXE) (eTable 1 in the Supplement). Extraocular extension was found in 56 eyes (1.7%) with a median diameter of 3 mm (IQR, 1.8-4.8 mm; range, 0.5-50 mm). According to prognostic groupings, 1030 patients (32.0%) had stage I disease, 1095 (34.0%) had stage IIA disease, 710 (22.1%) had stage IIB disease, 282 (8.8%) had stage IIIA disease, 79 (2.5%) had stage IIIB disease, and 21 (0.7%) had stage IIIC disease (Table 1).
Metastasis developed in 325 patients (10.1%) with ciliary body and choroidal melanoma. According to the AJCC criteria, of the 325 patients who developed metastasis, 27 (8.3%) had stage I disease, 85 (26.2%) had stage IIA disease, 111 (34.2%) had stage IIB disease, 69 (21.2%) had stage IIIA disease, 24 (7.4%) had stage IIIB disease, and 9 (2.8%) had stage IIIC disease. With the use of univariate analysis, metastasis was related to prognostic and anatomical staging (log-rank test for trend P < .001) (Table 2 and Figure 1).
Kaplan-Meier point estimates for all patients with primary ciliary body and choroidal melanoma who were metastasis free were 98% at 1 year, 86% at 5 years, and 81% at 10 years (Table 2). According to stage, the 5-year metastasis-free point estimates were 97% (95% CI, 95%-98%) for stage I, 89% (95% CI, 86%-91%) for stage IIA, 79% (95% CI, 75%-83%) for stage IIB, 67% (95% CI, 59%-73%) for stage IIIA, 50% (95% CI, 33%-65%) for stage IIB, and 25% (95% CI, 4%-53%) for stage IIIC. The 10-year metastasis-free point estimates were 94% (95% CI, 91%-96%) for stage I, 84% (95% CI, 80%-88%) for stage IIA, 70% (95% CI, 62%-76%) for stage IIB, 60% for stage IIIA (95% CI, 51%-68%), and 50% (95% CI, 33%-65%) for stage IIIB. Estimates for stage IIIC could not be calculated because of the small sample size (Table 2).
With the use of a Cox regression model, stage IIA and higher stages all have a significantly higher risk of metastasis compared with stage I (all hazard ratios [HRs] >1.0 and all P < .001) (eTable 2 in the Supplement). Compared with stage I, the HR was 3.1 for stage IIA, 6.6 for stage IIB, 12.2 for stage IIIA, 19.5 for stage IIIB, and 24.4 for stage IIIC. When each stage was compared with the stage before it, the risk of metastasis increased 3-fold for stage IIA (P < .001), 2-fold for stages IIB (P < .001) and IIIA (P < .001), and 1.6-fold for stage IIIB (P = .047). In stage IIIC (which only had 21 patients), the HR is 1.3 (compared with stage IIIB); however, this result was not statistically significant (P = .57).
Patients with primary ciliary body and choroidal melanoma classified according to tumor size category were as follows: T1, 1116 patients; T2, 1151 patients; T3, 776 patients; and T4, 174 patients (eTable 3 in the Supplement and Table 3). The numbers of patients with metastasis were as follows: T1, 30 (2.7%); T2, 117 (10.2%); T3, 135 (17.4%); and T4, 43 (24.7%).
Univariate analysis was used to determine Kaplan-Meier metastasis-free point estimates at 1, 5, and 10 years according to tumor size (eTable 3 in the Supplement, Table 3, and Figure 2A). Increasing tumor size was associated with an increased risk of metastasis (log-rank test for trend P < .001). Metastasis-free point estimates decreased at both 5 and 10 years were as follows: 97% (95% CI, 95%-98%) and 94% (95% CI, 91%-96%) for T1, 85% (95% CI, 82%-88%) and 80% (95% CI, 75%-84%) for T2, 77% (95% CI, 73%-80%) and 68% (95% CI, 60%-74%) for T3, and 61% (95% CI, 49%-71%) (5-year only) for T4, respectively (Table 3 and Figure 2A).
With the use of a Cox regression model, larger tumors all have a significantly higher risk of metastasis compared with T1 tumors (all HRs >1.0) (eTable 2 in the Supplement). Compared with T1 tumors, the HRs were 4.11 for T2, 7.46 for T3, and 14.45 for T4 (all P < .001). When each tumor size category was compared with the category before it, the risk of metastasis increased 4-fold for T2 (P < .001) and 2-fold for T3 (P < .001) and T4 (P < .001) tumors.
Patients with ciliary body and choroidal melanoma were then subclassified according to the presence or absence of CBI and/or EXE. Subcategories divided according to the AJCC manual were as follows: category a, no CBI or EXE (n = 2733); category b, CBI only (n = 428); category c, EXE only (n = 25); and category d, CBI and EXE (n = 31). Univariate analysis suggests that metastasis-free probability was related to CBI and/or EXE grouping (log-rank test for trend P < .001). Metastasis-free point estimates as shown in Kaplan-Meier curves were generally worst among patients with both CBI and EXE, followed by EXE only, then those with CBI only, and best among those with neither CBI nor EXE. The 5- and 10-year metastasis-free point estimates were 90% (95% CI, 88%-91%) and 84% (95% CI, 81%-86%) for no CBI or EXE, 72% (95% CI, 66%-77%) and 67% (95% CI, 60%-73%) for CBI only, 54% (95% CI, 29%-74%) (5-year only) for EXE only, and 33% (95% CI, 13%-54%) and 33% (95% CI, 13%-54%) for both CBI and EXE, respectively (Table 3 and Figure 2B).
Patients with ciliary body and choroidal melanoma were also subclassified according to size of the EXE as follows: no EXE, 3161 (98.3%); EXE of 5.0 mm or less, 44 (1.4%); and EXE greater than 5.0 mm, 12 (0.4%). The 5- and 10-year Kaplan-Meier metastasis-free point estimates were 87% (95% CI, 86%-89%) and 81% (95% CI, 79%-84%) for no EXE, 43% (95% CI, 25%-60%) and 43% (95% CI, 25%-60%) for EXE 5.0 mm or smaller, and 42% (95% CI, 7%-75%) (5-year only) for EXE greater than 5.0 mm, respectively (eTable 4 in the Supplement). The Kaplan-Meier plot suggests that metastasis is related to the size of the EXE (log-rank test for trend P < .001) (eFigure in the Supplement).
The median follow-up period for 160 patients with primary iris melanoma was 3.6 years (IQR, 1.6-6.4 years; range, 8.0-12.3 years). Of these tumors, 103 (64.4%) were biopsy proven. Tumors involved the ciliary body in 83 patients (51.9%) and the choroid in 6 patients (3.8%). Median tumor thickness was 3.0 mm (IQR, 2.0-5.0 mm; range, 0.5-9.0 mm). Secondary glaucoma was found in 40 patients (25.0%), scleral extension in 7 patients (4.4%), and extraocular extension in 2 (1.3%). For the 160 iris melanomas, tumor size categories there were as follows: T1, 1 (0.6%); T1a, 33 (20.6%); T1b, 26 (16.3%); T1c, 18 (11.3%); T2, 66 (41.3%); T2a, 14 (8.8%); T4, 1 (0.6%); and T4a, 1 (0.6%).
Five (3.1%) of 160 patients with iris melanoma developed metastasis; all had T2 tumors (by definition, extending into the ciliary body and/or choroid). Kaplan-Meier point estimates could not be calculated because of small sample size.
Our generation stands at the forefront of a new medical information age.24 With the use of a multicenter, international, Internet-based registry, the staging system for posterior uveal melanoma in the 7th edition of the AJCC Cancer Staging Manual has now been supported. The resultant number of patients was large enough to obtain statistically significant results. Specifically, we confirm that both higher tumor size category and prognostic stage are significantly related to increased risk of metastasis. The independent factors of CBI, EXE, and EXE size were also associated with an increased risk of metastasis, although the range of EXE size varied greatly (0.5-50 mm; eTable 1 in the Supplement).
The distribution of patients with choroidal and ciliary body melanoma by AJCC stage in the present study was comparable to the EOOG study.5 However, the present study had a higher percentage of patients who had stage I disease (32.0% vs 20.7%) and a lower percentage who had stage IIIA disease (8.8% vs 16.9%) and stage IIIB disease (2.5% vs 7.1%). Using the AJCC staging system, we found that the risk of metastasis at 10 years increased approximately 2-fold by stage (3.1 vs 6.6 vs 12.0 vs 19.5 vs 24.4 for stages IIA, IIB, IIIA, IIIB, and IIIC, as compared to stage I, respectively). These hazard ratios as well as our Kaplan-Meier metastasis-free point estimates for stages I to IIIC at 5 and 10 years after initial staging were comparable to those in the EOOG study.5 Although the median follow-up times in the present study and the EOOG study were comparable (3.2 vs 3.4 years), the larger difference observed in the 10-year rates may have been influenced by the fact that the EOOG included patients accrued during a wider range of years. In contrast, we chose to limit our recruitment window to patients diagnosed as having cancer from April 1, 2001, to April 1, 2011. This may be (in part) why the EOOG analysis included 1301 tumor deaths (18% of all patients), whereas in the present study only 325 patients (10.1%) developed metastasis. Most important, now 3 multicenter studies5,25,26 (EOOG, COMS, and OOTF) and 1 single-center study10 have reported that tumor size (largest diameter and thickness) is a clinical predictor of mortality from uveal melanoma.5,10,25,26
One prior multicenter, international, Internet-based study8 allowed collaboration among ophthalmic oncology centers. These authors examined 131 patients with biopsy-proven iris melanomas and found increasing mortality with increasing AJCC tumor category. In contrast, in our study only 64.4% of the iris melanomas were biopsy proven, and the mortality rate was too low to calculate statistical differences by tumor size category. This is further evidence that large multicenter studies are needed for these rare diseases and that future versions of the AJCC Cancer Staging Manual should consider updating the iris melanoma section.
The OOTF recognizes that future modifications of the AJCC staging system are inevitable. Future modifications are likely to involve incorporation of a patient’s genetic and molecular uveal melanoma characteristics. In our study, data were collected on chromosomal alterations and other biomarker features. However, we did not present those results because they are beyond the scope of this study.27- 29 We also expect additional studies in which AJCC tumor staging is used to define best clinical practices, treatment adverse effects, and evaluation of biomarkers. Validation of the TNM staging system is important as a baseline for future clinical trials of new agents that may offer a cure for metastatic uveal melanoma.
A weakness of our study is its retrospective design. For example, there was no opportunity to standardize methods of diagnosis, treatment, follow-up, or screening for metastatic disease. In addition, 432 (11.3%) of our 3809 total patient entries could not be used because of missing or inconsistent data.
This study found that retrospective, multicenter, international data sharing is feasible in ophthalmic oncology and can be used to evaluate tumor staging systems. In doing so, this study provides the framework for prospective standardized data sharing, which will, in turn, offer the best opportunity to evolve cancer-staging systems. Multicenter patient recruitment should decrease the effect of single-center selection bias and provide a more heterogeneous patient study population.
Submitted for Publication: August 26, 2014; final revision received October 27, 2014; accepted October 29, 2014.
Corresponding Author: Paul T. Finger, MD, The New York Eye Cancer Center, Ocular Tumor Service, 115 E 61st St, Fifth Floor, New York, NY 10065 (firstname.lastname@example.org).
Published Online: January 2, 2015. doi:10.1001/jamaophthalmol.2014.5395.
The AJCC Ophthalmic Oncology Task Force Member Authors (in order of numbers of patients contributed to the study) includes E. Rand Simpson, MD; Brenda L. Gallie, MD; Svetlana Saakyan, MD; Anush Amiryan, MD; Paul T. Finger, MD; Kimberly J. Chin, OD; Stefan Seregard, MD; Maria Fili, MD; Matthew Wilson, MD; Barrett Haik, MD; Jose M. Caminal, MD; Jaume Catala, MD; David E. Pelayes, MD, PhD; Anibal M. Folgar, MD; Martine Jager, MD; Mehmet Dogrusöz, MD; Arun Singh, MD; Andrew Schachat, MD; Shigenobu Suzuki, MD; Yukiko Aihara, MD.
Affiliations of The AJCC Ophthalmic Oncology Task Force Member Authors: Department of Ocular Oncology, Princess Margaret Cancer Centre, Toronto, Ontario, Canada (Simpson); Departments of Molecular Medical Genetics, Medical Biophysics, and Ophthalmology, Toronto, Ontario, Canada (Gallie); Department of Ophthalmic Oncology and Radiology, Helmholtz Institute, Moscow, Russia (Saakyan, Amiryan); Department of Ophthalmic Oncology, The New York Eye Cancer Center, New York, NY (Finger, Chin); Department of Ophthalmic Oncology, St. Erik’s Eye Hospital, Stockholm, Sweden (Seregard, Fili); The Ophthalmic Oncology Service, University of Tennessee, Memphis (Wilson, Haik); Department of Ophthalmology, Hospitalet de Llobregat, Barcelona, Spain (Caminal, Catala); Department of Ophthalmic Oncology, Buenos Aires University, Buenos Aires, Argentina (Pelayes, Folgar); Department of Ophthalmic Oncology, Academisch Ziekenhuis Leiden, Leiden, the Netherlands (Jager, Dogrusöz); Department of Ophthalmic Oncology, Cole Eye Institute, Cleveland, Ohio (Singh, Schachat); Division of Ophthalmic Oncology, National Cancer Center Hospital, Tokyo, Japan (Suzuki, Aihara).
Author Contributions: Dr Finger had had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.
Study concept and design: Simpson, Gallie, Finger.
Acquisition, analysis, or interpretation of data: All authors.
Drafting of the manuscript: Finger, Chin.
Critical revision of the manuscript for important intellectual content: All authors.
Statistical analysis: Finger, Chin.
Obtained funding: Finger.
Administrative, technical, or material support: All authors.
Study supervision: Simpson, Gallie, Saakyan, Finger, Seregard, Wilson, Caminal, Pelayes, Jager, Singh, Suzuki.
Conflict of Interest Disclosures: All authors have completed and submitted the ICMJE Form for Disclosure of Potential Conflicts of Interest and none were reported.
Funding/Support: This project was supported by The Eye Cancer Foundation, the AJCC, and the Paul T. Finger, MD, Fund at Princess Margaret Cancer Centre.
Role of the Funder/Sponsor: The funding source had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and the decision to submit the manuscript for publication.
Additional Contributions: Princess Margaret Cancer Centre, Toronto, Ontario, Canada, served as the coordinating center and provided statistical analysis through Christine Massey, MS; Priya Durairaj, MD, provided extensive data collection and oversaw overall data quality; Yuliya Gavrylyuk, MD, and the Health Informatics Research team developed the algorithms and database and managed the Internet website. The Institut Catala Oncologia acknowledges Cristina Gutierrez, MD, Department of Radiotherapy, Juan Pera, MD, Department of Radiotherapy, and Josep M. Piulats, MD, Department of Medical Oncology, for their participation in uveal melanoma treatment and research. Helmholtz Institute, Moscow, acknowledges Vladimir Valskiy, MD, PhD, for his contribution in uveal melanoma treatment. The New York Eye Cancer Center acknowledges Ekaterina Semenova, MD, for her contribution to data collection. The Leiden University Medical Center acknowledges Stefan de Geus, MD, and G. P. M. Luyten, MD, PhD, for their contributions. Staging forms are reproduced with the permission of the AJCC. The original source for this material is the 7th edition of the AJCC Cancer Staging Manual. Ms Massey was compensated for her contribution to statistical analysis. No other individuals were compensated.
Correction: This article was corrected on March 9, 2015, to fix an error in the byline and on July 30, 2015, to fix an error in the Discussion section.