(PDF 578 KB)
Eye cancer specialists have long been handicapped by a fundamental communication gap: Unlike other areas of oncology, ophthalmic oncology has been slow to adopt a common language for categorizing ocular tumors.
How, then, can ophthalmologists compare outcomes for various modalities or from various centers if tumors are grouped in a hodgepodge of ways? Until recently, with the exception of a few well-designed studies, it was difficult to do so with any reliability or reproducibility. But that may now change, with the help of the updated classification and staging system from the combined American Joint Committee on Cancer (AJCC) and International Union for Cancer Control (UICC) Ophthalmic Oncology Task Force. The recently published AJCC Cancer Staging Manual1 offers several improvements over earlier editions—including a more complete classification of ocular tumors, the addition of biomarkers and inclusion of guidelines for data acquisition—that promise to make them more relevant to ophthalmic researchers and clinicians.
Yet one barrier remains: widespread adoption of this system by the ophthalmic community.
Why AJCC Adoption Is Important
The AJCC criteria are based on tumor size (T), regional nodal status (N) and distant metastasis (M). “Universal adoption of the TNM nomenclature can move ophthalmic oncology into the mainstream of medical oncology,” said Paul T. Finger, MD, chairman of the AJCC-UICC Ophthalmic Oncology Task Force and clinical professor of ophthalmology at New York University School of Medicine.
Consistent staging clarifies research. Bita Esmaeli, MD, professor of ophthalmology at the University of Texas M. D. Anderson Cancer Center, agreed that widespread use of the AJCC criteria is important to the profession: “If you read the ophthalmic oncology literature, with the exception of perhaps retinoblastoma and uveal melanoma, reports on eye cancers such as eyelid carcinoma, conjunctival melanomas, lacrimal gland cancers and sarcomas of the orbit don’t even begin to categorize the cancers at presentation and elucidate what stage they’re in. Investigators report outcomes without paying attention to the stage of disease at presentation. So we end up comparing apples and oranges—a mixture of stages bunched together.
“For example, some of the most quoted, highly regarded papers on adenoid cystic carcinoma of the lacrimal gland—the key studies that suggest improved survival associated with certain treatments—provide no information about size and stage of tumors at presentation or predominant histologic type. But these are all important factors that may make a difference to outcomes,” Dr. Esmaeli said. Indeed, a recent study by Dr. Esmaeli and colleagues found that T classification (size of the eyelid tumor) at presentation for sebaceous carcinoma correlates with the risk of regional lymph node metastasis and death from this disease.2
|These eyelid tumor presentations would typically have been grouped together in research studies without regard to categorization. According to the AJCC 7th edition, these tumors would be classified (top) T2b and (bottom) T3a.
What’s New in This Edition?
Although previous editions of the AJCC Cancer Staging Manual have been available for years, the publication of the 7th edition has energized eye cancer specialists to use AJCC criteria for outcomes studies. There are several reasons why.
Created by the ophthalmic community. First, previous editions were produced by small committees, while the 7th edition recruited a large number of ocular tumor specialists to create it. The task force of 45 subspecialists from 11 countries included experts in ophthalmic pathology, ocular plastic surgery, ocular oncology, and orbital and adnexal tumors. “A concerted effort was made to ensure that the AJCC criteria represent the community,” Dr. Finger said, adding, “What’s made by the community will be used by the community!”
More complete and balanced. The new edition expands coverage to tumors omitted in earlier editions, for example, ocular adnexal lymphomas. And it streamlines classification for diseases such as retinoblastoma, for which multiple systems were used, which made it difficult for researchers and clinicians to draw accurate conclusions.
Biomarkers have been added. The new edition now incorporates biomarkers. “We don’t have a lot of biomarkers in ophthalmic oncology,” said Dr. Finger, “but we have some. And we’re looking at new ones, which can be incorporated in future editions. For example, the standard uptake values (SUVs) of PET scans may have prognostic value for the ability of a choroidal melanoma to metastasize. We are collecting SUVs for these tumors whenever possible.”
Includes guidelines for data acquisition. Most eye cancers are rare, making it particularly important to employ a uniform classification when reporting treatment outcomes. Dr. Finger said that, in general, reports have been based on “single-institution experiences that are relatively small.” And even at larger centers, data have been collected according to each institution’s preferences and methods, he said.
The new edition addresses that problem by providing guidelines on data acquisition. “We’re hoping to get everyone collecting data in the same way so that it’s possible to add these data together. To get any kind of statistical significance, you typically need hundreds of patients; but for rare cancers, any single institution might be able to collect data on only 20 cases. It’s not that those findings aren’t valid; it’s just that they can’t achieve statistical significance. There is power in numbers!”
Helps harness the power of numbers through data sharing. Dr. Finger said that it’s much less expensive to share data than it is to do prospective, randomized trials. “And the number of cases we could include would be far greater. For example, even though the Collaborative Ocular Melanoma Study was a tribute to cooperation in collecting a total of 2,500 melanomas, that number is only what is seen in one year in North America. And the study took 16 years to complete (though data acquisition was completed in a shorter time). If we had a way of adding up multiple centers’ clinical experiences, we would be able to achieve statistical significance similar to that of prospective, randomized trials faster.”
Implementing AJCC Criteria
“The barrier to universal adoption of the AJCC criteria is simply that most ophthalmologists are not in the habit of using them,” Dr. Finger said. “We’re doing everything we can to ensure that the current criteria will be implemented.”
Getting journals on board. “We’ve shown the editors of the major journals that the updated criteria are truly representative of the subspecialty,” said Dr. Finger. In fact, 12 major peer-reviewed journals now include the use of AJCC criteria as part of their instructions for authors. And, according to Dr. Finger, hospitals with tumor registries already use them.
Checklist for clinics. To facilitate staging of each new orbital, eyelid and conjunctival tumor prospectively, the ophthalmology clinics and operating rooms at the M. D. Anderson Cancer Center have been using a one-page intake sheet adapted from the new AJCC manual. Dr. Esmaeli made the intake sheet available at http://links.lww.com/IOP/A50.3 “Of course we don’t deem the sheet perfect, but it can be used easily in a busy setting and can become part of electronic medical records.”
Validating AJCC Criteria
“Several recent studies4,5 support the value and clinical relevance of a staging system, but any staging system has to be validated through repeated and widespread application,” said Dr. Esmaeli. “There is a critical need to publish outcome data using these criteria. This represents a tremendous opportunity for investigators.”
Internet validation surveys. Dr. Finger said, “This is the best effort of 45 people from 11 countries, but we are realists and recognize that we’re working with very little evidence-based data. That’s why the AJCC-UICC Ophthalmic Oncology Task Force has embarked on a huge endeavor to validate all the chapters of the 7th edition. We are creating Internet-based validation surveys to collect significant numbers retrospectively. We want to test how well the AJCC criteria work. Whatever flaws we find, we will rectify them in the 8th edition.
“One validation survey (for choroidal melanoma) is live already. We’re looking to put it into 10 to 20 centers, with each contributing about 500 patients, so that we’ll have at least 5,000 patients to test. Retinoblastoma and conjunctival melanoma surveys are currently under construction. We’re hoping to get about 1,000 cases for rarer tumors, such as lacrimal gland cancer, where the largest series in the literature is about 60.” Dr. Finger said that being able to look back and quantify eye cancers and treatments “offers us an avenue to elevate the quality of care for these tumors.”
Developing an ASOPRS database. Dr. Esmaeli is leading another validation effort, sponsored by the American Society of Ophthalmic Plastic and Reconstructive Surgery (ASOPRS) Foundation, to set up a multi-institution database using AJCC criteria for all the tumors that the society’s members treat: tumors of the eyelid, conjunctiva and orbit, as well as uveal melanoma and retinoblastoma.
“We envision adding several thousand new patients a year to the database,” said Dr. Esmaeli. “We’re going to capture data prospectively and, when possible, retrospectively for cases that are well documented enough to give us the TNM status.” A paper Dr. Esmaeli and colleagues published on the feasibility of using AJCC criteria for eyelid carcinomas shows that it’s possible to collect data even retrospectively, though there will be some gaps in any retrospective attempt.6 She emphasized that, of course, it’s always better to collect TNM information prospectively when a patient is first evaluated.
Ultimate goals: saving vision and lives. All of these efforts are putting ophthalmic oncology in good stead for closing its communication gap. “A common language is the foundation from which we can evolve to sharing data and, by sharing data, figure out what diagnostic and therapeutic techniques are working for patients, ultimately saving vision and saving lives,” said Dr. Finger.
Dr. Esmaeli is a recipient of an ASOPRS Foundation grant for the set-up of a multi-institutional ocular oncology database. Dr. Finger has no financial interest in the AJCC or clinical staging. However, his research and part of the expenses for AJCC work (some calls, meetings, etc.) were paid for by the Eye Cancer Foundation, for which he serves as executive director.
1 Edge S, ed. AJCC Cancer Staging Manual. 7th ed. New York: Springer; 2010.
2 Esmaeli B et al. Ophthalmology. In press.
3 Droll L et al. Ophthal Plast Reconstr Surg. 2011;27(2):142.
4 Ahmad SM et al. Ophthalmology. 2009;116(6):1210-1215.
5 Damato B, Coupland SE. Clin Experiment Ophthalmol. 2008;36(8):786-795.
6 Shinder R et al. Orbit. 2011;30(5):202-207.
FURTHER READING: See the online edition of this article for suggested additional reading.