This article is more than 5 years old. Information may no longer be current.
Small breast cancers may be overdiagnosed because they are ‘fundamentally different’
Click Here to Manage Email Alerts
Many small breast cancers may not grow large enough to pose a threat during a women’s lifetime and could lead to overdiagnosis and unnecessary treatment, according to a special report in The New England Journal of Medicine.
“Breast cancers are not all the same and there are major differences in the growth rate and behavior from one tumor to the next,” Donald R. Lannin, MD, professor of surgery at Yale School of Medicine, told HemOnc Today. “This paper shows that a large portion of small cancers are small, not because they were diagnosed early, but because they are biologically favorable and grow so slowly that they will never become large. Therefore, small cancers have a good prognosis, not because they were caught earlier, but because they are fundamentally different in their composition.”
The expected rate of overdiagnosis in women with breast cancer is 22%, according to a 2016 study by Welch and colleagues published in The New England Journal of Medicine.
Lannin and his colleague, Shiyi Wang, MD, assistant professor of epidemiology at Yale Public School of Health, used SEER data to analyze invasive breast cancers diagnosed between 2001 and 2013, which they divided them into three prognostic groups based on grade, ER status and PR status. Researchers defined each as favorable, intermediate or unfavorable and used the expected rate of overdiagnosis (22%) to categorize the types of breast cancers and patient ages that likely account for the majority of overdiagnoses.
Researchers hypothesized the favorable tumor group (grade 1 ER, PR positive; grade 1 ER positive, PR negative; and grade 1 ER negative, PR positive) would have the highest rate of overdiagnosis and the unfavorable tumor group (grade 2 ER, PR negative; grade 3 ER, PR negative; grade 3 ER positive, PR negative; and grade 3 ER negative, PR positive) would have the lowest rate of overdiagnosis, while the intermediate tumor group would fall in between.
Overdiagnosis findings
Lannin and Wang reported that among women aged 40 years or older, tumors with favorable biologic features represented up to 38.2% of tumors smaller than 1 cm but only 9% of tumors larger than 5 cm. Tumors with unfavorable biologic features comprised 14.1% of smaller tumors and 35.8% of larger tumors.
These findings appeared similar in women aged younger than 40 years, but favorable tumors were only half as common and unfavorable tumors occurred more commonly.
Researchers then created three models with varied distributions of favorability.
Model 1 assumed 41% of diagnosed cancers are favorable, 55% intermediate and 4% unfavorable. Under this model with a 22% rate of overdiagnosis, researchers found 40% of favorable cancers, 22.3% of intermediate cancers and 3.8% of unfavorable cancers are overdiagnosed.
The mean lead time — or time between when a cancer can be detected by screening and when it would have become clinically apparent without screening — was 19.9 years for favorable tumors, 10.6 years for intermediate tumors and 2 years for unfavorable tumors.
Model 2 assumed 53% of cancers diagnosed were favorable, 44% intermediate and 3% unfavorable. Under this model, 51.8% of favorable tumors were overdiagnosed with a mean lead time of 29.6 years. Only 2.9% of unfavorable cancers were overdiagnosed with a mean lead time of 1.4 years.
Researchers used this model to show the percentage of overdiagnosis according to age and tumor biologic features. Under the assumption that the mean lead time varies according to biologic group and not patient age, “the model clearly shows that overdiagnosis is less common in young women and increases steadily with age,” the researchers wrote.
Model 3 — which assumed 65% of cancer are favorable, 33% intermediate and 2% unfavorable — showed overdiagnosis ranging from 63.5% with a mean lead time of 44.9 years for favorable tumors to 2.1% with a mean lead time of 0.9 years for unfavorable tumors.
Researchers also used these three models under overall overdiagnosis rates of 16.5% and 11%. Although lead times under these rates varied widely, “all the models showed that the lead time for the favorable tumors was at least an order of magnitude greater than that of the unfavorable tumors,” the researchers wrote.
Implications for screening
In its 2016 recommendation, the U.S. Preventive Services Task Force (USPSTF) states that for women who are at average risk for breast cancer, most of the benefit of mammography results from biennial screening for women aged 50 to 74 years.
Women aged 60 to 69 years are most likely to avoid breast cancer death through mammography screening, according to the USPSTF recommendation.
The recommendation notes screening women aged 40 to 49 years may reduce breast cancer death, but at a smaller rate than in the older age groups and at the expense of false-positive results and unnecessary biopsy.
Although a large portion of aggressive cancers grow quickly — with a lead time of 2 years or less — many other breast cancers grow so slowly that the lead time is 15 to 20 years, meaning a woman aged 75 years or older diagnosed with a small, slow-growing tumor may opt for active surveillance, Lannin said.
“Medical groups should consider possibly changing guidelines about when to stop mammography as the risks for overdiagnosis are so high in older women,” Lannin said. “This study may help justify less aggressive treatment, such as foregoing radiation or sentinel node biopsy in women over 70 years and not giving chemotherapy to low-risk oncotype patients.”
Moreover, these data suggest that although mammography is capable of diagnosing small tumors, many of those cancers may never grow large enough during the lifespan of the patient to be diagnosed without mammography.
“The paper provides some models that, for the first time, suggest which types of cancers and patient age ranges likely account for the majority of overdiagnosis,” Lannin said. “In addition, those favorable cancers that do increase in size still have a very good prognosis, so there is little value in finding them early.
“In contrast, large cancers that contribute to the greatest portion of breast cancer deaths get large so quickly that often they cannot be diagnosed by mammography before they become palpable,” Lannin added. “This paper explains, therefore, both how mammography causes overdiagnosis and also why it is not more effective. More importantly, it questions some of our fundamental beliefs about the value of early detection.”
The American Cancer Society recommends yearly mammograms for women aged 45 to 54 years, and biennial screenings for women 55 years and older, continuing as long as a woman is in good health and expects to live at least 10 more years.
These data suggest that younger women are more likely to develop aggressive cancers than older women, thus reinforcing the need for annual mammography screening starting at age 40 years, according to a press release issued by the American College of Radiology.
“Aggressive cancers in women in their 40s must be identified early to be effectively treated,” Debra Monticello, MD, FACR, chair of the America College of Radiology Breast Imaging Commission, said in the release. “Letting these tumors grow even 1 extra year before screening greatly increases odds that the woman will die from breast cancer.”
In a 2012 study by Otto et al, nearly 30% of unscreened women who died of breast cancer between 1995 and 2003 developed stage IV tumors, compared with 5% of women who had been screened.
“Giving the second-leading cancer killer in women a head start through reduced or delayed screening can be lethal for these women,” Wendy B. DeMartini, MD, president of the Society of Breast Imaging, said in the release. “Starting screening at age 40 remains the best policy.”– by Chuck Gormley
References:
Welch HG, et al. N Engl J Med 2016;doi: 10.1056/NEJMoa1600249
Otto SJ, et al. Cancer Epidemiol Biomarkers Prev. 2012;doi:10.1158/1055-9965.EPI-11-0476.
For more information:
Donald R. Lannin , MD, can be reached at the Department of Surgery, Yale University School of Medicine, P.O. Box 208062, New Haven, CT 06520; email: donald.lannin@yale.edu.
Disclosure: Lannin and Wang report no relevant financial disclosures.
Click Here to Manage Email Alerts