Cancer Research Highlights
Decades of Data Point to Overdiagnosis from Breast Cancer Screening
Since breast cancer screening came into widespread use in the United States in the 1970s, more than 1 million women may have been diagnosed with cancers that never would have caused them harm or required treatment, a new study suggests. These women may have been exposed unnecessarily to the adverse effects of treatment, the authors reported in the November 22 New England Journal of Medicine.
The detection of cancers that do not grow or grow so slowly that they would never cause illness is known as overdiagnosis. Previous studies have shown that screening mammography, which looks for breast cancer in the absence of symptoms, can lead to overdiagnosis.
Overdiagnosis may account for nearly one-third of newly diagnosed breast cancers among women aged 40 and older in the United States, the authors of the new study estimated. In 2008, for example, more than 70,000 women may have received an unnecessary diagnosis, they noted.
"This is a significant public health concern," said co-author Dr. Archie Bleyer of St. Charles Health System in Bend, OR. "Women need to be aware of the potential benefits of screening, as well as the downsides—including being diagnosed with cancers that [are not life-threatening]."
To look for evidence of overdiagnosis, Dr. Bleyer and Dr. H. Gilbert Welch of Dartmouth Medical School in Hanover, NH, used NCI's Surveillance, Epidemiology, and End Results (SEER) database to analyze trends in breast cancer incidence between 1976 and 2008.
The authors reasoned that if screening leads to the earlier detection of cancers that are destined to become lethal, detecting more breast cancers at an earlier stage—when they tend to be curable—should lead to a corresponding drop in late-stage cancers. But the SEER data did not show this to be true: The rise in early-stage breast cancers over three decades (an absolute increase of 122 cases per 100,000 women) was not matched by an equivalent drop in late-stage cancers. Instead, there was an absolute decrease of 8 cases per 100,000 women. This imbalance, the authors concluded, must be due to overdiagnosis.
This estimate of overdiagnosis is generally consistent with estimates from other countries. A recent Norwegian study found that as many as 1 in 4 invasive breast cancers diagnosed in that country through its population-based mammography screening program never would have caused harm.
Nonetheless, comparing studies can be a challenge because of differences in study design. For instance, Drs. Bleyer and Welch counted noninvasive tumors known as ductal carcinomas in situ among the early-stage breast cancers, whereas the Norwegian researchers did not.
A limitation of the current study was the fact that the authors had to infer overdiagnosis from incidence statistics in the population, because overdiagnosis cannot be directly observed at the individual patient level.
The study does not clarify whether an individual woman should be screened for breast cancer, the authors acknowledged. But they noted that the potential harms of unnecessary diagnoses are clear: emotional stress and anxiety, surgery, radiation therapy, hormonal therapy, chemotherapy, or, as is often the case, a combination of these treatments—all for abnormalities that would not have caused illness.
“Women need to understand that screening has positive and negative consequences,” said Dr. Stephen Taplin of NCI’s Division of Cancer Control and Population Sciences, who has studied screening for 25 years but was not involved in this study. “But they also need to know that a decision about screening is not a forever choice. A woman can choose to be screened later, or not at all.”
He added, “This is one of many studies that is expanding the discussion about screening. It demonstrates that women need to make decisions based on their circumstances, not just based on recommendations.”
After Negative Colonoscopy, Rescreening with Other Tests May Be Effective
According to a new modeling study, people who have a colonoscopy that finds no precancerous polyps (a negative colonoscopy) at age 50 can be rescreened beginning at age 60 with one of three alternative methods rather than having colonoscopies every 10 years, without affecting their life expectancy. Rescreening with one of the alternative methods—highly sensitive fecal occult blood testing (HSFOBT), fecal immunochemical testing (FIT), or computed tomographic colonography (CTC or "virtual colonoscopy")—would also cause fewer complications and cost less.
These results, from the NCI-funded Cancer Intervention and Surveillance Modeling Network (CISNET) team from the University of Minnesota School of Public Health and their colleagues, were published November 6 in the Annals of Internal Medicine.
Most current guidelines recommend rescreening with colonoscopy 10 years after an initial negative colonoscopy. However, these recommendations are not based on results from randomized trials. "There are ongoing trials of colonoscopy, but none of them have reported results yet," said lead author Dr. Amy Knudsen.
The researchers used a model called SimCRC, which was used to inform the 2008 update of the United States Preventive Services Task Force guidelines on colorectal cancer screening. Dr. Knudsen and her colleagues used the model to simulate five different rescreening strategies: no further screening, colonoscopy every 10 years, HSFOBT every year, FIT every year, or CTC every 5 years.
Two adherence scenarios were evaluated: one in which people received the tests as scheduled (perfect adherence) and one that mimicked real-life adherence, as recorded in several published studies (imperfect adherence).
The results were the same in both scenarios: all four rescreening methods reduced the number of deaths from colorectal cancer compared with no rescreening, and the difference among the four methods was small. For example, the imperfect-adherence scenario yielded between 6.1 and 6.7 deaths per 1,000 persons for all four screening methods. (See the table.)
Rescreening with colonoscopy not only produced the highest rates of perforation (tears in the colon) and other complications, but it was the most expensive strategy. Rescreening with one of the other three screening methods produced lifetime savings of up to $495 per person, compared with imperfect adherence with colonoscopy. (See the table.) At a population level, these savings could add up to nearly $3 billion for HSFOBT or FIT, and $0.6 billion for CTC, over the lifetimes of the estimated 6.5 million people who had negative colonoscopy results in 2008.
"Models can be helpful to inform [population] guidelines overall. On an individual level, decisions should be made in consultation with one's doctor," concluded Dr. Knudsen.
Deaths per 1,000 People
Estimated Lifetime Savings Per Person, Compared with Colonoscopy
|Fecal immunochemical testing|
|Computed tomographic colonography|
|Highly sensitive fecal occult blood test|
This study was funded by the National Institutes of Health (RC1CA147256 and grants U01CA088204, U01CA152959.)
Longer Delays in Breast Cancer Treatment May Affect Survival
Delays between a breast cancer diagnosis and treatment increase the risk of death for women with late-stage cancers, according to a study published November 19 in the Journal of Clinical Oncology (JCO). A second study appearing in the same issue of JCO found that the median wait time between diagnosis and treatment has grown longer. According to the authors, findings from the studies may provide data that can be used to develop quality measures for breast cancer care.
The first study looked at the length of time between a breast cancer diagnosis and the start of treatment among women who were enrolled in the North Carolina Medicaid system. Dr. John M. McLaughlin and his colleagues found that women with late-stage breast cancer who waited more than 60 days between diagnosis and treatment had a 66 percent greater risk of death from any cause and an 85 percent greater risk of death from breast cancer than women who began treatment within 60 days.
By contrast, treatment delays of more than 60 days in patients with early-stage breast cancer were not associated with survival differences.
The findings "suggest that interventions should target late-stage patients to increase the timeliness of receiving breast cancer treatments and that clinicians should structure their practice settings to promptly triage and initiate treatment for patients diagnosed at late stage," the authors wrote.
The second study looked at patients with nonmetastatic breast cancer from the SEER-Medicare Linked Database and found that the median wait between the first physician visit and first surgery rose from 21 days in 1992 to 32 days in 2005. Dr. Richard Bleicher of the Fox Chase Cancer Center in Philadelphia and his colleagues found that times to surgery were longest among black and Hispanic patients, patients in the northeast, and patients in large metropolitan areas. The researchers also found that more complex surgeries—for example, a simultaneous mastectomy and reconstruction—were associated with longer wait times.
Imaging, biopsies, and clinician visits all made statistically significant contributions to surgery delays. "More episodes of care may cause delay but may allow for better assessment of treatment alternatives," the authors explained.
"One challenge in caring for patients is not just to give quality care but to give timely care," Dr. Bleicher said. "My hope is that this study provides physicians—and patients—a point of reference for time to surgery that we've never really had before."
These studies were funded in part by the National Institutes of Health (1R01CA121317 and N01-PC-35136).
Fusion Gene Linked to Rare Form of Childhood Leukemia
Researchers have identified a fusion gene that may drive some rare and difficult-to-treat cancers in children called acute megakaryoblastic leukemias (AMKL). The investigators have also developed a test that can detect this genetic change in children at the time of diagnosis, which will help doctors identify candidates for future clinical trials of much-needed new therapies.
The findings, published November 13 in Cancer Cell, are from the Pediatric Cancer Genome Project. Led by St. Jude Children's Research Hospital in Memphis and Washington University in St. Louis, the project is analyzing the normal and cancer genomes of hundreds of children and adolescents with a variety of cancers.
This particular study focused on AMKL that develops in children without Down syndrome, an aggressive form of the disease whose biology is poorly understood. (Children with Down syndrome who develop AMKL have excellent prognoses.)
The researchers initially found the fusion gene in 7 of the first 14 patients whose RNA they analyzed, which was an unexpectedly high frequency. After finding the fusion gene in a substantial fraction of another, larger group of patients, the researchers estimated that 27 percent of children with AMKL have the fusion gene.
When the investigators looked at survival statistics, the results were striking. At St. Jude, only 34 percent of children with the fusion gene were alive 5 years after diagnosis, compared with nearly 89 percent of those who lacked the fusion gene. AMKL in children without Down syndrome "is very rare, but the outcomes are very poor," noted Dr. Tanja Gruber of St. Jude, the study's first author.
The fusion gene codes for a chimeric protein that includes part of CBFA2T3, a protein that helps immature blood cells continue to divide (proliferate), and GLIS2, a transcription factor that is expressed in the kidneys and that had not been associated previously with cancer. Studies in mice suggested that the fusion protein helps immature blood cells proliferate for longer than normal cells.
In further experiments, the authors traced this effect to a signaling pathway known as BMP, which was much more active in cells with the fusion gene than in cells without it. The fusion protein is likely to affect other signaling pathways as well, the researchers said.
No drugs are currently available to block the effects of the fusion protein. But a diagnostic test based on a laboratory technique called PCR will enable clinical trials to begin as promising treatments emerge, Dr. Gruber noted. For now, the fusion gene could be a marker of poor prognosis.
The researchers also found that patients with the fusion gene had, on average, seven other genetic changes, whereas patients without the fusion had 17 other genetic changes. "This tells us that while you do need additional mutations beyond the fusion gene to cause the disease, you need very few of them," said Dr. Gruber. Which of these lesions are important in the disease still needs to be determined in future studies, she added.
This work was supported in part by the National Institutes of Health (P30 CA021765).
Alternative Type of Brachytherapy Proves Effective in Mice
An injectable genetically engineered peptide polymer may one day offer an alternative to conventional brachytherapy, a commonly used radiation therapy technique, according to the results of a new study in mice. The treatment could eliminate some of the difficulties associated with brachytherapy, the researchers believe, and could be used to treat more cancer types than brachytherapy. The study results were published November 15 in Cancer Research.
Using mouse models of two different cancer types, the researchers showed that their alternative brachytherapy approach—directly injecting tumors with a biodegradable elastin-like polypeptide (ELP) that is "labeled" with radioactive iodine—effectively shrank tumors and, in many cases, eliminated them completely.
In conventional brachytherapy, radioactive seeds are implanted in tumors and later removed. This type of internal radiation therapy is used frequently to treat localized prostate cancer and, to a lesser extent, to treat breast cancer. The seeds must be implanted and removed surgically, however. Another disadvantage is that implanted seeds can travel to other, healthy tissues, explained the study's lead investigator, Dr. Wenge Liu of Duke University, and his colleagues.
By contrast, ELPs are liquid at room temperature and can be injected into tumors. Once inside tumors, they assemble into small seeds, or depots. In the study, the researchers tested ELP formulations that varied in their amino acid composition, size, and concentration to determine which one developed the most stable depots and was retained longest in the tumor before breaking down.
The researchers identified the ELP that had demonstrated the best tumor retention, and they tested three variations in mouse models of prostate and head and neck cancers. After only a single administration, the most effective variation shrank tumors in all of the mice regardless of tumor type. It also completely eliminated tumors in two-thirds of the mice with head-and-neck tumors and all of the mice with prostate tumors.
A potential advantage of the ELP is that it eventually breaks down into nontoxic forms that are naturally excreted by the body, Dr. Liu and his colleagues wrote. They reported no clinical signs of side effects in the treated mice, and additional examinations showed that the radioactive iodine was concentrated at the tumor site, with very little accumulation in healthy tissues.
In addition to treating several types of localized tumors, an ELP has other potential uses, the researchers wrote, including shrinking (debulking) tumors considered to be inoperable so that they can be removed surgically.
The researchers are continuing to refine the approach, Dr. Liu said in an e-mail message, including investigating whether the radioactive iodine dose can be lowered without sacrificing efficacy and working to improve delivery of the ELP "to solid tumors located deep in the body, such as in the esophagus, bronchus, stomach, and colon or abdominal cavity."
This research was supported in part by the National Institutes of Health (5R01CA138784-03).