Study: Breast cancers can disappear without treatment: fact or fiction?

IN-DEPTH REVIEW OF RESEARCH

Study background:

Screening mammography has been successful. The death rate from breast cancer was previously unchanged for decades before the introduction of routine mammography in the U.S. in the mid-1980s. Since then, death from breast cancer has decreased by 37% in recent decades for women over 40. This decrease is largely credited to routine breast screening by mammograms. However, previous research and accompanying news articles have claimed that some breast cancers detected by routine screening can disappear or regress on their own without treatment.

Elizabeth Arleo and her colleagues from New York-Presbyterian/Weill Cornell Imaging and other institutions published a research article in the Journal of the American College of Radiology in July 2017 which reports results of a prospective study designed to understand what happens to untreated breast cancers.

Researchers of this study wanted to:

Learn what happens to untreated breast cancers detected through routine screening. Do they spontaneously disappear or regress?

Population(s) looked at in the study:

The researchers sent a 14 question survey by email to all 108 active fellows of the Society of Breast Imagining in clinical practice; 42 fellows responded to this survey and contributed data. Fellows were asked about:

• the number of screening mammograms interpreted in their practice in the last year and the last 10 years 
• the number of screening-detected breast cancers in their practice in the last year and last 10 years 
• whether any of the screening-detected breast cancers went untreated. The survey instructed doctors to consider cancer as “untreated” when, after a was biopsy performed, patients with pathology-confirmed breast cancer received no surgery, radiation, chemotherapy, hormonal or other treatment. If doctors reported that some of their patients’ breast cancers were untreated, they were sent a follow-up questionnaire that included questions about cancer “regression,” for example, whether the cancer “remained visible at subsequent mammography, but no longer appeared suspicious for malignancy.”  

Study findings:​ 

• The total number of screening mammograms among all practices was 6,865,324.
• The total number of screening-detected invasive breast cancers was 25,281.
  • Of these, 240 cases were not treated.
    • None of the untreated cases of invasive breast cancers was reported to spontaneously disappear or regress on subsequent mammography screening.
• The total number of screening-detected cases of ductal carcinoma in situ was 9,360.
  • Of these, 239 cases were not treated.
    • None of the untreated cases of ductal carcinoma in situ was reported to spontaneously disappear or regress on subsequent mammography screening.

The most controversial harm attributed to routine breast cancer screening is overdiagnosis. Though the study authors note that an unknown percentage of breast cancers detected by routine mammography in this study may represent overdiagnosis, all of the untreated cancers were still visible at the next screening mammogram. Because they observed no breast cancers that spontaneously disappeared or regressed, the authors suggest that their findings support the argument that overdiagnosis should not be a factor when considering when to begin and how often to screen for breast cancer. They argue that delaying the onset of screening or increasing the interval between screenings will not reduce the frequency of overdiagnosis. For example, they argue, if a woman begins routine screening at age 40 or 45, an overdiagnosed, non-cancer lesion may be detected and possibly result in overtreatment. However, if this same woman begins routine screening at age 50, that same overdiagnosed non-cancer lesion will still be there and may still result in overtreatment.

Limitations:

This research study relies on survey data which has many limitations. Some questions relied on the radiologist's memory which can lead to incorrect responses. Because the survey was anonymous, it is not known who responded to the survey, or where they practiced. Additionally, the authors of this research study do not know why some of the patients diagnosed with breast cancer were not treated. Importantly, individual patient records were not reviewed. Thus, there is no way to determine if patients received any additional intervention which limits the ability to draw meaningful conclusions from this type of study. Finally, there is no data on patient characteristics or their outcomes.

The authors state that “Our results show that all untreated, screen-detected cancers persist” and imply that this is always the case.  It is difficult to conclude this based solely on the results of this survey. This study was a survey based on analysis of 6,865,324 mammograms over a 2-week period of time.  Without reviewing individual patient’s records for the 479 untreated cases, one cannot assume that none of these patients received additional interventions. Further, the authors suggest that “starting screening at a later age … and increasing the screening interval … will simply delay the timing of overdiagnosis but will not decrease the frequency of overdiagnosis.” While this may be true, it is not a conclusion that can be made based on the study design. Finally, to conclude that the survey proves that “overdiagnosis is not mitigated by less screening” is similarly not supported by this report.         

Conclusions:

This study suggests that prior studies and myths that some breast cancers detected during routine screening can regress or spontaneously disappear without treatment are not correct. The authors state that their findings do not challenge the existence of overdiagnosis; rather, they show that all untreated screening-detected cancers persist and will be interpreted as suspicious at the next screening mammography. These conclusions should be interpreted keeping in mind the many limitations of this study. Women should work with their health care providers to determine the screening schedule that is best for them, taking into account their personal and familial history of cancer.

Share your thoughts on this XRAYS article by taking our brief survey.