Preface

Established in 1973, the Surveillance, Epidemiology, and End Results (SEER) Program of the National Cancer Institute (NCI) has been a valued source of high-quality information for cancer incidence and survival in the United States. SEER data are used in a variety of research studies to explore cancer incidence and survival for etiologic and outcomes research related to cancer. These studies include exploring factors that influence trends, analyzing and understanding the cancer burden at a national and local level, and describing health disparities among subpopulations. Data are collected on approximately 450 000 cases of malignant and in situ cancers each year, including information on patient demographics, primary tumor site, tumor morphology and stage at diagnosis, first course of treatment, and mortality outcomes. The registry community strives to capture clinically relevant information by periodically expanding the scope of data collected, such as incorporating the latest staging definitions, as well as predictive and prognostic factors as they become part of standard care. Staging describes the extent of an individual's cancer at the time of diagnosis and plays a vital role in managing patient care. Cancer staging reflects prognosis and is used in determining appropriate treatment for cancer patients. It is an essential component in many areas of cancer research, from defining eligibility criteria in clinical trials to evaluating the impact of cancer interventions and treatment advances in epidemiologic and health services research. The American Joint Committee on Cancer (AJCC) leads the effort in the United States to develop a standardized staging system and collaborates with the Union for International Cancer Control (UICC) to maintain a system that is used worldwide. Cancer staging guidelines as well as classification systems for tumors continue to evolve over time and are updated periodically to reflect advances in cancer care and diagnosis. The AJCC and UICC work closely with cancer registrars to ensure that registries can capture the most current staging definitions. The SEER Program and its predecessors such as the End Results Program have collected information to stage cancers and have adapted their data collection to evolving staging requirements (Fig. 1). Staging over time. Traditionally, staging represented anatomic information regarding the scope of the disease, and this information was summarized into 3 categories: the extent to which the tumor has invaded nearby tissue (T; including size of the tumor for some sites), whether the cancer has spread to nearby lymph nodes (N), and whether the cancer has metastasized to other parts of the body (M). Together, T, N, and M form the basis for cancer staging. With advances in medical technologies and the introduction of precision medicine and targeted therapies, staging systems incorporated factors that influence prognosis and treatment decisions to reflect current medical practice. As the complexity of the components that determine prognosis continued to increase, it became increasingly challenging to maintain staging systems that would both provide consistency over time and capture the factors influencing treatment decisions in the current medical setting. To address the challenge, a set of building blocks was developed that would provide the foundation needed to translate the extent of disease information into multiple staging systems and enhance the anatomic staging information with additional clinically relevant factors such as biomarkers and prognostic factors. This new approach resulted in the creation of the Collaborative Stage Data Collection System (CS). The first version of Collaborative Stage (CSv1) was introduced for cases diagnosed in 2004 and coincided with the release of the 6th edition of the AJCC Cancer Staging Manual.1 CS is a unified data collection system designed to use a single set of data elements based on extent of disease and clinically relevant factors that both meets the needs of multiple staging systems and eliminates duplicate data collection by cancer registrars reporting to facility-based and central population-based registries. The project was sponsored by the AJCC in collaboration with NCI's SEER Program, the American College of Surgeons Commission on Cancer (CoC), the Centers for Disease Control and Prevention's National Program of Cancer Registries, the National Cancer Registrars Association, the North American Association of Central Cancer Registries, and the Canadian Cancer Society's National Cancer Institute of Canada. The Collaborative Staging system was updated in 2010 (CSv2) in conjunction with the release of the 7th edition of the AJCC Cancer Staging Manual (AJCC 7th)2. The authors of the AJCC 7th present the T, N, M, and stage descriptions along with a section on prognostic factors for which collection is recommended. These recommendations are the basis for the development and collection of the site-specific factors (SSFs) for CS. NCI and the CoC decided to require many, but not all, of these new clinically significant SSFs for their respective programs, SEER and the National Cancer Data Base. These additional data elements have great potential for expanding our understanding of patient diagnosis, stage, and outcomes. Additional information contained in the SSFs allows for the identification of more homogeneous groups of patients who share common disease characteristics beyond anatomic similarities at the cancer sites. CSv2 was designed to allow computer algorithms to calculate multiple staging systems, simultaneously giving the most current staging and preserving the ability to analyze longer-term trends by stage. The identification and analysis of comparable stage-specific cancer incidence trends over time are essential for assessing the impact of cancer-control interventions and for comparative effectiveness research, thus playing a vital role in cancer surveillance. This issue looks at 8 common cancer sites to describe the information collected under CSv1 and CSv2 within the SEER Program. Data analyzed are from 18 SEER registries that represent 28% of the US population. Each report first describes how changes between the AJCC 6th and 7th affect stage distributions and trends and then quantifies the potential impacts on outcomes and incidence trends stratified by stage. SSFs are described in detail with particular emphasis on the factors newly collected in 2010. To better evaluate the completeness of each SSF, 2 separate analyses were performed. The first analysis considered all cases for which the information was collected, whereas the second was restricted to the cases in which the information would be expected to be routinely found. For example, if a SSF is included in existing guidelines to determine treatment, it would be expected to be routinely found. Key questions related to the SSFs are addressed in the discussion sections of each article, such as how a factor relates to other information collected and the completeness and quality of the information. The level of completeness of SSF information for cases diagnosed in 2010 ranged from very high (over 95%) to very low. It will require additional investigation to understand the reasons for some of the low percentages of completeness. Some test results may be missed because the test was given outside the reporting facility; it is also possible that a test was not done for any number of reasons not routinely captured in the registry data, such as comorbidities or individual preference. In several instances, there were differences between laboratory values and test interpretation, which called into question whether pooling laboratory values across different laboratories without additional information associated with the test performed could be trusted to provide consistent and accurate information. Considerable resources have been spent on the collection of these data, and they should be available for research when the data are of high quality, consistent, and complete (low percentage unknown). Analyzing similar SSF data for 15 major cancers, NCI formed a Data Release Work Group composed of SEER principal investigators, SEER registry staff, and NCI staff to make recommendations on SSFs that should be released to researchers and SSFs that should no longer be collected as of 2014. The evaluation for the remaining sites/schemas continues. Additional information on issues that researchers need to understand in analyzing stage, the CS building blocks of stage (extent of disease), and the CS SSFs can be found on the SEER website: http://surveillance.cancer.gov/reports/. The aim of this report is to provide information and interpretations that will serve as a guide to users of SEER registry data. There are many ongoing efforts that will facilitate the collection of predictive and prognostic factors in the future. Central cancer registries obtaining information directly from laboratories and physician offices through the increased use of electronic pathology reports and electronic medical records will lead to more complete and consistent reporting. More defined data structure formats within electronic health records may allow for easier electronic data capture and reporting of these items in the future, removing some of the burden from the cancer registrar. Linkages to other supporting data sets, such as medical claims and pharmaceutical transaction databases, may be useful for filling in missing information. In the future, as the College of American Pathology checklists (http://www.cap.org/apps/cap.portal) and AJCC are more closely aligned, information on SSFs and basic staging information may be more readily available to cancer registrars. New and creative approaches to capturing data will continue to be needed to keep up with the ongoing advances in understanding, describing, and treating cancer. Even as AJCC 7th stage data are analyzed, groups are already meeting to design the AJCC 8th edition in order to maintain its relevance and pivotal role in patient care. This special issue was made possible by NCI's Surveillance Research Program, experts from SEER registries and other leaders from the surveillance community. These individuals spent many hours reviewing the data that were collected under CSv1 and CSv2 and are dedicated to ensuring that the data are used to their fullest potential by the research community. This study would not be possible without the dedication of the staff at the SEER population-based cancer registries and the staff at the facilities who provide data to the SEER central registries. This supplement edition of Cancer has been sponsored by the National Cancer Institute. Data used in the production of this supplement was supported under Contract HHSN261201300004I (University of Southern California), HHSN261201300005I (Cancer Prevention Institute of California, HHSN261201300009I (University of Hawaii), HHSN261201300010I (University of New Mexico), HHSN261201300021I (Rutgers University), HHSN261201300019I (Connecticut Department of Health), HHSN261201300020I (University of Iowa), HHSN261201300015I (Emory University), HHSN261201300016I (Louisiana State University), HHSN261201300017I (University of Utah), HHSN261201300011I (Wayne State University), HHSN261201300012I (Fred Hutchinson Cancer Center), HHSN261201300013O (University of Kentucky), and HHSN261201300014I (Public Health Institute). Technical support was provided under contract HHSN261201100007I (Information Management Services, Inc.). Dr. Ries was supported by contracts HHSN261201300308P and HHSN261201200422P from the National Cancer Institute for work related to the current study. The authors made no disclosures.