Collecting clues: Cancer registries might have an answer
■ The data gathered in these repositories play a key role in the constant struggle against cancer morbidity and mortality.
By Kathleen Phalen Tomaselli amednews correspondent — Posted April 17, 2006
- WITH THIS STORY:
- » Evolution of cancer registries
- » External links
- » Related content
Elyce H. Cardonick, MD, didn't begin her cancer registry with the first pregnant cancer patient she treated. It was after she saw several and was struck by the scarcity of related case information that she started the collection.
"The last thing people think about is walking into an oncologist's office pregnant," says Dr. Cardonick, assistant professor of obstetrics and gynecology in the division of maternal fetal medicine at Cooper University Hospital in Hamilton, N.J. She now maintains her own cancer registry of 240 women from around the world.
"Approaches are so individualized. Collecting cases can lead to a more standardized way of caring for these women. With the registry, I can say, 'Here are 80 others with breast cancer who were pregnant, and this was their treatment; here are their outcomes.' ... It gives the doctor and the patient confidence."
Dr. Cardonick's effort is but one example of how cancer registries function on a variety of levels -- locally, nationally and internationally. Overall, these intertwined repositories are an important part of efforts to reach a better understanding of diagnosis, treatment and prevention.
The concept behind them has been in use for centuries. Records indicate that the earliest observations of patterns of disease occurred in the 17th century, as elders gathered death information for publication in London's weekly Bill of Mortality. By the 18th century, Bernardino Ramazzini, in his De Morbis Artificum Diatriba, noted an increased risk of breast cancer in nuns, and Sir Percival Potts linked scrotal cancer in chimney sweeps to soot exposure.
Through the years, the evolution of modern cancer registries continued. In 1956, the American College of Surgeons developed an approval program for hospital-based cancer registries. Today, 75% of all national data are submitted to the ACS.
"The value of cancer registries in population research is immeasurable," says Holly L. Howe, PhD, director of the North American Assn. of Central Cancer Registries, a collaborative organization for cancer registries, governmental agencies and professional and private groups. "Through linkage with other data systems, the information can give us insight into getting effective treatments to the general population that will have an impact on survival and mortality."
Early registries were little more than uncultivated card files of data. Not until government funding supported state and national registry programs and laws establishing many cancers as reportable diseases did these caches of information become vital weapons in the war against this disease.
"Cancer registries are key in tracking the U.S. cancer burden. They provide a real foundation for research," says Brenda Edwards, PhD, associate director for the National Cancer Institute's Surveillance, Epidemiology and End Results program, or SEER. "We've been collecting data since 1973, tracking incidence, survival, treatment patterns by modality, treatment-related complications, patterns of care by location or population, cost of care, survival and treatment by race, and quality of life of survivors."
Primary care physicians play an important role in this system, because they are on the front lines of seeing patients and families, says James Evans, MD, PhD, a geneticist and director of the adult genetics program at the University of North Carolina in Chapel Hill, which keeps a genetic cancer registry. "They are the most important link in all this."
Data bubbling upwards
These repositories track life, death, cancer diagnosis and treatment, and often operate in seamless silence as data increase year by year. And as scientists study meticulously gathered patient stories, cancer's mysterious language begins to unravel, and public health interventions become possible.
"The data is bubbling from the bottom," says Sharon Metzger, the manager of the cancer registry at Johns Hopkins Kimmel Cancer Center in Baltimore. "We can look at cancer across the board -- local, regional, state, national. We gather benchmarking data to really improve patient care and quality."
Examples of how registries lead to improvements in care and detection of patterns dot the national and international landscape.
In South Carolina, for instance, Karen Mace's daily work with a hospital registry begins with a stack of charts from which she gleans patient data -- demographic, medical history, co-morbidities, type of cancer, diagnosis, stage at diagnosis and treatment. It's an archive she'll update for each patient entered in the registry for the remainder of that patient's life. It initially can take a registrar more than an hour per case.
"We collect the data from the medical record," says Mace, a certified tumor cancer registrar for Bon Secours St. Francis Health System in Greenville, S.C., adding that the Health Insurance Portability and Accountability Act requires releases that all patients sign when admitted. "Then I call physicians and get any missing information."
As part of the process, daily admissions are checked for patients in the registry. In addition, the Social Security Death Index is scanned, and letters are sent to primary care doctors, surgeons and oncologists.
"In the annual letters, we ask physicians, 'Have you seen or treated the patient, are they disease-free, are they alive, did they have a recurrence?' Because we are an American College of Surgeons-approved cancer center, we must have 90% follow-up each year," Mace says.
Like Mace, Johns Hopkins registrars collect data on newly diagnosed and readmitted patients. With 5,600 cases a year, it takes six registrars to complete collection for local use, quarterly state reporting and annual national uploads. While some hospital registries remain underutilized, the Hopkins' registry is active with requests for data extraction for administrative planning, outcomes analysis and cancer prevalence, trends and lengths of stay, to name a few.
"We get over 100 requests a year from just inside Johns Hopkins," Metzger says. "In the past it was difficult, because data were collected manually and a lot of data could not be retrieved. But in the past 20 years that's changed, and now we have the ability to return data quickly for tons of requests."
State and national registries
Mace manages 650 cases for the 344-bed Greenville hospital. In the same city, Greenville Memorial manages 1,000 cases. And the data collected from both are funneled into a central state registry of 20,000. These patient cancer stories are compiled again and sent annually to both the SEER registry of more than 6 million and the ACS registry, which sets cancer program and registry standards and has data on more than 15 million cases. States report data to NCI's SEER, the ACS Registry, the CDC National Program of Cancer Registries and the National Cancer Database.
In 1998 and 1999, the South Carolina registry, along with the state's Dept. of Public Health, identified a true cancer cluster of malignant mesothelioma -- cancer of the lining of the lung -- in Charleston. "The cluster was directly caused by asbestos and was easily tracked to shipbuilding," says Susan Bolick-Aldrich, director of the South Carolina Central Cancer Registry in Columbia. "The community requested an assessment. And we rallied round this with stepped-up screenings and community meetings about the risks. We're still following the numbers."
Meanwhile, across the registry spectrum, tight rules assure consistent reporting among registrars and reporting bodies. All use identical staging and reporting manuals, and approved programs must have a certified registrar to assure consistency. "It is this uniformity of pooled data that makes things work," Dr. Edwards says. "This couldn't be done without collaboration. The key to all this is the hospital registrars and what the hospital does. We all have to work together."
Dr. Edwards admits there is room for improvement. For instance, the lag time from collection to reporting is 24 months. "Right now, 2004 data will be published at the end of the year," Bolick-Aldrich says. "We've improved in the last 10 years, but it needs more improvement. With new technology and more streamlining of data collection, we can do it. This is a very exciting time for cancer registries."
Applying the insights
Back in 1996, Dr. Cardonick met two pregnant women with cancer: one with Hodgkin's disease, the other with melanoma. "The first was referred to me for a termination of her pregnancy," Dr. Cardonick says. "But she did not want the termination, so I had to ask, 'Will the termination make her cancer better?' "
She began researching the literature and found that some patients had been safely treated while pregnant. Her patient ended up receiving chemotherapy and doing well.
After seeing more and more cases, Dr. Cardonick started her registry to track treatments and outcomes. "I get their medical record to see what was done. I get information on drugs and doses. I also get consent forms for the patient's oncologist, OB, pediatrician, and every year on the baby's birthday, I send a letter to find out how mom and baby are doing."
Dr. Cardonick's archive is unique, as are familial and genetic registries that track families with high concentrations of certain cancers. Because genetic and molecular markers are important to these registries, the human genome project greatly accelerated their work.
"The hope is that by looking at a whole subset of cancers we can gain more insight," Dr. Evans says. "We have learned a great deal about colorectal cancer by studying these relatively unusual families. By looking at enough patients, we can detect patterns and associations."
The UNC registry, which has a familial and genetic cancer registry of 3,000 patients, is one of eight in the NCI's Cancer Genetics Network. Dr. Evans says the network is at the forefront of a national trend toward individualized medicine. Specializing in the study of inherited predisposition to cancer, its expanding database has information on 24,000 individuals and 16,000 families, with a four-generation history.
"This population of people can help tell us about molecular defects and pathways of cancer," says Joyce Seldon, a medical geneticist at the UCLA Familial Cancer Registry and Genetic Evaluation Program at the Johnson Comprehensive Cancer Center in Los Angeles, which has a registry of 800. In a cassette-like 2-inch-by-1-inch wax block, the tissue is banked and saved indefinitely. "It is very valuable. There may already be molecular alterations even without cancer."
But, similar to other types of registries, keeping up with familial and genetic registries is much more than tissue and blood banking and includes quality of life and psychosocial issues. It's a little like following a moving target. "We track after one month, six months, 12 months, then for life," Seldon says.