Cancer Research Highlights
Study Finds Colonoscopy Screening Overused in Medicare Beneficiaries
A new study suggests that almost one-quarter of Medicare beneficiaries undergo colorectal cancer screening with colonoscopy more frequently than is recommended, including a large number of beneficiaries 80 years of age or older. Performing potentially unnecessary colonoscopies in such elderly beneficiaries is of "special concern" because the risks associated with the procedure (such as intestinal perforations and infections) can often outweigh the benefits for this age group, the study authors noted.
Dr. James Goodwin of the University of Texas Medical Branch, Galveston, and his colleagues published their findings online May 9 in the Archives of Internal Medicine.
Previous studies have documented the underuse of colorectal cancer screening in general, but this study examined whether colonoscopy was being overused in certain populations based on current recommendations. After a negative result on a screening colonoscopy, cancer organizations and public health groups almost uniformly recommend that most people not have a repeat screening colonoscopy for another 10 years.
For this recent study, the researchers analyzed claims and enrollment data from a random sample of Medicare beneficiaries between 2000 and 2008. The analysis included adjustments to identify initial and repeat screening colonoscopies, since these procedures often are not coded as such on Medicare claims, the study authors noted.
Approximately 46 percent of the more than 24,000 Medicare beneficiaries who had a negative result on the initial screening colonoscopy between 2001 and 2003 underwent another colonoscopy sometime in the next 7 years. In this group, nearly 43 percent had "no clear indication for the early repeated examination," the researchers found, suggesting that the colonoscopies were screening procedures.
A third of people who were 80 years of age or older at their initial negative screening had what appeared to be another screening colonoscopy within 7 years. The analysis revealed "inflection points with noticeable increases in the rate of repeated colonoscopies at 3 and 5 years, suggesting that those procedures might have been routinely scheduled rather than in response to symptoms."
Limiting inappropriate or medically unnecessary colonoscopies is important for a number of reasons, the researchers wrote. Apart from exposing patients to unnecessary risks and incurring extra costs, they continued, unnecessary colonoscopies tie up resources that could otherwise be used "to increase appropriate colonoscopy in inadequately screened populations."
Prostate Cancer Study Provides More Data on Surgery versus Watchful Waiting
Surgery may be the preferred option for men younger than age 65 who are diagnosed with early-stage prostate cancer, according to extended follow-up from a randomized clinical trial. For several reasons, however, the findings may have limited applicability to men in the United States who are diagnosed with early-stage disease, several researchers said.
The results were published in the May 5 New England Journal of Medicine.
The clinical trial, conducted in Sweden, Finland, and Iceland, randomly assigned 695 men newly diagnosed with early-stage prostate cancer to immediate surgical removal of the prostate (radical prostatectomy) or watchful waiting, a less-intensive precursor to what today is typically called active surveillance. Among men younger than age 65, those who underwent immediate surgery had superior overall survival and prostate cancer-specific survival, and their risk of dying from prostate cancer was 51 percent lower than those in the watchful waiting group. (In the 15-year follow-up period, among men younger than 65, 28 men in the surgery group and 49 men in the watchful waiting group died from prostate cancer. The numbers of prostate cancer deaths for all men in the two groups were 55 and 81, respectively.)
Overall, about 15 men had to be treated to prevent one death, reported Dr. Anna Bill-Axelson of the University Hospital, Uppsala, and her colleagues. For men younger than 65, however, the number who needed to be treated to avoid one death was about seven.
The trial "has provided important evidence that effective treatment is both necessary and possible for many men with early-stage prostate cancer," wrote Dr. Matthew R. Smith of the Massachusetts General Hospital Cancer Center in an accompanying editorial. But the findings, he cautioned, "may not be relevant for men with low-risk early-stage prostate cancers identified by PSA screening." In the United States, the vast majority of early-stage prostate cancers are diagnosed by PSA screening, whereas only about 5 percent of cases in the Swedish trial were diagnosed in this manner.
Surgery and other common prostate cancer treatments can have serious side effects, including incontinence and erectile dysfunction, explained Dr. Barry Kramer, editor-in-chief of NCI's Physician Data Query Screening and Prevention Editorial Board. "Because PSA screening can lead to the detection of a large number of tumors that wouldn't have caused any medical problems in the absence of screening," he continued, "the ratio of benefit to risk for immediate surgery can be quite different for men with PSA-detected cancers."
In addition, the watchful waiting protocol used in the trial differs significantly from the more aggressive active surveillance protocol often used in the United States. Men in the watchful waiting arm of the trial were considered for treatment (with transurethral resection of the prostate) only if they had specific symptoms that suggested their disease was progressing.
In contrast, active surveillance programs in the United States can entail twice-yearly checkups that may include PSA testing and digital rectal exams, as well as annual or regular prostate biopsies. These may be followed by curative treatment with surgery or radiation if there is evidence of disease progression.
A recent observational study found that prostate cancer-specific survival rates in men with early-stage disease diagnosed via PSA screening who opted for active surveillance were nearly identical to those of men who opted for immediate treatment.
Results from a large, randomized U.S. trial, called PIVOT, comparing surgery with watchful waiting in men diagnosed with localized prostate cancer, are expected to be available sometime this year. Because the large majority of the men in the trial were diagnosed by PSA screening, the findings should be more applicable to current U.S. practice, Dr. Kramer said.
Breast Cancers Arising between Mammograms Have Aggressive Features
Breast cancers that are discovered in the period between regular screening mammograms—known as interval cancers—are more likely to have features associated with aggressive behavior and a poor prognosis than cancers found via screening mammograms. These high-risk features include higher stage and grade, larger size, and lack of estrogen receptors (ER) and progesterone receptors (PR), researchers reported online May 3 in the Journal of the National Cancer Institute.
The results highlight the need for more sensitive screening methods and for women to monitor their breast health between mammograms, concluded Dr. Victoria Kirsh of Cancer Care Ontario and her colleagues.
The researchers compared the traits of breast cancers in women who had been screened in the Ontario Breast Screening Program between 1994 and 2002. The analysis included 87 women who had "missed" interval cancers, 288 women who had true interval cancers, and 450 women whose cancers were detected by screening mammography. Missed interval cancers were cancers that could have been detected by mammography but were overlooked due to error or difficulty in reading the x-ray films. True interval cancers were not detectible at the time of last screening, even on review.
Some clear differences between the tumors were found. Compared with screen-detected cancers, true interval cancers were almost four times as likely to be larger than 2 cm in diameter, more than four times as likely to be stage III or IV instead of stage I, more than three times as likely to be poorly differentiated (to have very abnormal-looking cells), about three times as likely to have a high proliferative rate (to be growing quickly), and about twice as likely to be estrogen receptor negative and progesterone receptor negative. Missed interval cancers were also more likely to be large, to be poorly differentiated, and to have invaded the lymph nodes than screen-detected cancers.
Women in the study were older than age 50 and predominantly white. It is not clear whether these results would be similar in younger women or in a more diverse population.
"A small number of cancers will not be detected by mammography," explained co-author Dr. Anna Chiarelli in an e-mail. This study, she continued, serves as a general reminder that "women should be 'breast aware' and see their health care provider if breast symptoms arise between mammograms."
Patient Reports of Family Cancer History Are Often Inaccurate
Getting an accurate family history of cancer from individuals is important for doctors who make recommendations about cancer screening and prevention strategies based, in part, on this information. But, until now, clinicians have lacked good evidence about the reliability of these histories. A study published online May 11 in the Journal of the National Cancer Institute shows that when people in the general population are asked about the history of specific cancer types in their family their responses are often inaccurate.
Dr. Phuong Mai of NCI's Division of Cancer Epidemiology and Genetics and her colleagues investigated the reliability of reported family histories for the four most common cancers in adults: breast, colorectal, lung, and prostate. Overall, the researchers found that reports of no family history of cancer were highly accurate, but the accuracy of reports of specific cancers among relatives was low to moderate and varied by cancer type. Accuracy was highest for breast cancer, lowest for colorectal cancer, and better overall for first-degree relatives than second-degree relatives.
An accurate family history is important because some cancer screening and prevention recommendations are based on a person's risk level, which is determined in part by family history, noted Dr. Mai. Inaccurate histories can lead to incorrect risk estimates, which may result in unnecessary screening for some people and not enough for others.
The study, begun by Dr. Louise Wideroff, who was part of NCI's Division of Cancer Control and Population Sciences until recently, used information from the 2001 Connecticut Family Health Study, a random telephone survey of households in Connecticut. In the survey, 1,019 Connecticut residents reported on the history of various cancers in a total of 20,578 first- and second-degree relatives.
The researchers then tried to confirm reported cancer cases for a randomly selected subset of 2,605 of those relatives using a number of data resources, including state cancer registries, Medicare databases, the National Death Index, death certificates, and health care facility records, as well as through direct interviews with the relatives or with proxies for those who had died.
"We would strongly encourage people to learn as much as possible about their family history [of disease], cancer or otherwise, and take initiative to collect and preserve these valuable records," said Dr. Mai. "And clinicians need to be aware that when patients report that cancers have occurred in their family, it might require additional validation."
Confirming diagnoses, however, can be time-consuming, expensive, and difficult. In the future such validation may be facilitated by the availability of electronic medical records, Dr. Mai noted.
Physicians need to approach patients' family history information "with healthy skepticism," wrote Drs. Rachel Freedman and Judy Garber of the Dana-Farber Cancer Institute in an accompanying editorial. "Although we should thoughtfully listen to our patients' histories, we must listen even harder to what they 'could' be telling us, especially when specific information could influence their care and the care of their relatives," they concluded.