Issue: July 10, 2015
July 10, 2015
18 min read

As DCIS incidence increases, ‘no easy answers’ available to guide patient care

Issue: July 10, 2015
You've successfully added to your alerts. You will receive an email when new content is published.

Click Here to Manage Email Alerts

We were unable to process your request. Please try again later. If you continue to have this issue please contact

Breast cancer is the most common malignancy among U.S. women, accounting for more than 220,000 diagnoses and nearly 41,000 deaths per year, according to CDC estimates.

The projections are even bleaker, as results of a modeling study show breast cancer incidence will increase by more than 50% in the next 15 years.

Despite the staggering statistics, an increasingly vocal segment of the clinical community contends the numbers are skewed by the skyrocketing incidence of in situ cancers.

The most common type of noninvasive breast cancer is ductal carcinoma in situ (DCIS), in which abnormal cells grow inside the milk ducts.

“The individual cells look like cancer cells — they look like cells that are growing and making new cells — but they are what I call ‘dumb cells,’” Benjamin D. Smith, MD, associate professor of radiation oncology at The University of Texas MD Anderson Cancer Center, told HemOnc Today. “If they remain trapped inside the duct of the breast, they are no threat to anyone.”

Many women with DCIS may be exposed to unnecessary treatment, according to Benjamin D. Smith, MD, of The University of Texas MD Anderson Cancer Center.

Many women with DCIS may be exposed to unnecessary treatment, according to Benjamin D. Smith, MD, of The University of Texas MD Anderson Cancer Center. “There is also a converse argument to be made that we are doing some good in identifying these DCIS cases and treating them at an earlier stage,” Smith said.

Source: Photo by William Guerra

If the abnormal cells spread to other tissue, they can become life-threatening. However, the rate at which this occurs has not been firmly established, and clinicians and researchers are limited in their abilities to predict which cases will progress.

Because 90% of DCIS cases are detected by mammography, some suggest increasingly powerful screening methods are fueling an epidemic of overdiagnosis, forcing patients with a noninvasive disease that poses no immediate threat to undergo costly and potentially harmful treatments. They even argue referring to DCIS as “cancer” triggers unnecessary fear.

Others contend the debate surrounding DCIS primarily is an academic one that ultimately may harm patients if they do not receive treatment upon diagnosis.

HemOnc Today spoke with several breast cancer experts about the controversy surrounding DCIS, the effect enhanced screening capabilities are having on diagnosis, the appropriate management for patients with the condition and whether DCIS should be called “cancer.”

Increased incidence

Incidence of DCIS increased from 5.8 per 100,000 women in the 1970s to 32.5 per 100,000 in the early 21st century, according to an overview of the condition published in an ASCO educational book in 2012.

DCIS now accounts for about 1 of every 4 new breast cancer cases in the United States, American Cancer Society statistics show.

It also will be a key driver in a dramatic increase in breast cancer incidence, according to study results published in Journal of the National Cancer Institute. Philip S. Rosenberg, PhD, senior investigator in the division of cancer epidemiology at the NCI, and colleagues conducted a modeling study that forecasts trends in breast cancer incidence in the United States.

“We became aware by reading the literature that in situ cancers have become a really big part of the total pool of breast cancer cases,” Rosenberg told HemOnc Today. “The time has really come to analyze in situ cancers. It hasn’t been part of most analytic studies historically, and I felt its inclusion was a natural next step for us.”

Rosenberg and colleagues used SEER data, Census Bureau population projections and mathematical forecasting models to forecast the change in breast cancer incidence between 2011 and 2030 among women aged 30 to 84 years.

They determined the total number of breast cancer cases will increase by more than 55% during that period, from 283,000 cases in 2011 to 441,000 cases in 2030.

Researchers projected the proportion of ER-positive invasive breast cancers will remain level, continuing to account for approximately 63% of cases. However, they projected ER-positive in situ cancers will account for a larger percentage of cases in 2030 than they did in 2011 (29% vs. 19%). ER-negative cancers — which include difficult-to-treat triple-negative and HER-2–positive breast cancer subtypes — will account for a smaller percentage of cases in 2030 than in 2011 (9% vs. 17%).


“Our results suggest that, although breast cancer overall is going to increase, different subtypes of breast cancer are moving in different directions and on different trajectories,” Rosenberg said. “These distinct patterns within the overall breast cancer picture highlight key research opportunities that could inform smarter screening and kinder, gentler and more effective treatment.”

Historically, 20% to 25% of diagnosed breast cancers are DCIS, according to Daniel B. Kopans, MD, professor of radiology at Harvard Medical School and director of breast imaging at Massachusetts General Hospital.

“There is no reason to expect this will increase,” Kopans told HemOnc Today. “Their estimates that the rate of DCIS will increase is not accurate unless more women participate in screening each year. Invasive cancers, on the other hand, have been increasing steadily by 1% per year since 1940 and there is no reason to expect that this will stop.”

The ‘overdiagnosis’ debate

The use of screening mammography is grounded on the premise that early detection allows for treatment of breast cancer in its earliest and most curable stages.

However, even those who see value in mammography — as well as newer technologies that allow for better evaluation of breast tissue — contend it creates a double-edged sword.

Archie Bleyer, MD, clinical research professor at the Knight Cancer Institute at Oregon Health & Science University, told HemOnc Today in 2013 that DCIS “hardly existed” prior to widespread use of mammography.

“No one is denying that screening mammography doesn’t have some benefit,” Bleyer said. “The problem is that screening mammography has also created this situation of the overdiagnosis of breast cancer. … Millions of women have gotten therapy who did not need it.”

Still, 40,000 women die each year from breast cancer, Kopans said.

“Were they all, therefore, ‘overtreated’?” Kopans said. “Invasive breast cancers will likely kill if given enough time. There is little, if any, ‘overdiagnosis’ of invasive breast cancer.”

Although DCIS can progress to invasive carcinoma, the rate at which this occurs is not clearly established. Previous studies suggest 14% to 60% of DCIS cases will progress to invasive breast cancer within 10 years, according to the ACS’s Cancer Facts & Figures 2015 report.

“It’s helpful to think of DCIS not as a single entity, but as a group of different types of pre-cancer that have different risks associated with them,” Smith said. “At one end of the spectrum, you might get someone with a very small, mammogram-detected lesion that may only measure a few millimeters, and that type of lesion is not very risky. On the other side of the spectrum, you might find a breast full of very high-grade, aggressive DCIS. That patient is at considerable risk for conversion to invasive cancer without appropriate therapy.”

The challenge for clinicians is that, aside from tumor grade, few tools exist to accurately predict risk for progression. Although the risk factors for DCIS and invasive breast cancer are similar, the role of prognostic factors such as HER-2, ER and PR status in DCIS have not been firmly established.

Hence the dilemma about whether all women diagnosed with DCIS should undergo treatment.

Todd M. Tuttle, MD, MS

Todd M. Tuttle

“I think we diagnose breast cancer in many women who will never experience any harm from the disease,” Todd M. Tuttle, MD, MS, chief of surgical oncology at the University of Minnesota Medical School and co-author of the DCIS review published by ASCO, told HemOnc Today. “For these patients, overdiagnosis may lead to overtreatment, such as unnecessary radiation, surgery and endocrine therapy.”

Beth A. Virnig, PhD, MPH, senior associate dean for academic affairs and research at University of Minnesota School of Public Health and co-author on the ASCO review, voiced a similar sentiment.


“You could see [in the data presented by Rosenberg and colleagues] that the percentage of cancers that are ER-negative is expected to go down, and those are very serious cancers,” Virnig told HemOnc Today. “It makes me wonder: If cancer incidence is going up but it’s getting less serious, it’s getting less serious because we are diagnosing an awful lot of stuff that shouldn’t be diagnosed. … We need to ask ourselves, ‘What are the chances the patient will actually benefit from treatment?’”

Others disagree.

“The myth has been created by flawed methodology that has been able to get past poor peer review,” Kopans said. “All of the papers that have claimed massive overdiagnosis have been shown to have major methodological failures. The false claims of ‘overdiagnosis’ have already resulted in declining access to breast cancer screening.”

The rate of invasive breast cancer increased in the mid 1980s and in 1999 as more women participated in screening, Kopans said.

“This is the well-known ‘prevalence peak’ that happens with every screening test,” he added. “Once the proportion of women participating in screening becomes constant, the annual detection rate will return to the prescreening levels. The fear that screening is going to be the cause of increasing incidence is simply not scientifically supported.”

The ACS’s guidelines for breast cancer screening recommend initiating mammography at age 40 years, whereas the U.S. Preventive Services Task Force (USPSTF) recommends women undergo biennial breast cancer screenings between ages 50 and 74 years.

The USPSTF concluded the potential harms of mammography outweigh the benefits for women aged 40 to 49 years.

“The USPSTF felt that overdiagnosis and false-positives outweighed the fact that lives would be lost from breast cancer, and insurance follows USPSTF guidelines,” Kopans said. “The public health concern is that women will die from breast cancer whose lives could have been saved by annual screening starting at age 40.”

Although the detection of disease at its earliest stage is a universal goal in oncology, the decision to pursue aggressive therapy immediately must be balanced against the potential benefit of that treatment, as well as the potential for treatment-related adverse effects.

“The earlier the stage we diagnose breast cancer, the more treatable it is and the less invasive and destructive the treatments are,” Virnig said. “Nobody wants to see someone go into chemotherapy if they can avoid it.

“We are getting very good at keeping people alive,” Virnig said. “I don’t think we want to say, ‘Because you can live with very serious disease, we should allow people to delay getting diagnosed.’ We want to avoid the very serious disease. But at the same time, we know that unnecessary surgery and the amount of time spent in health care also present quality-of-life problems.”

Treatment options

Once clinicians decide to proceed with treatment for patients diagnosed with DCIS, the question becomes: Which approach is most appropriate?

The most common option is breast-conserving surgery plus radiation. Unilateral or bilateral mastectomies sometimes are recommended for patients with more aggressive cases. Endocrine therapy also is occasionally used, particularly for patients with ER-positive tumors.

“Radiation therapy and endocrine treatment can reduce the risk of a subsequent invasive breast cancer,” Tuttle said. “But these treatments also have risks.”

Most clinicians agree that treatment should be personalized, suggesting that disease- and patient-specific factors should guide treatment decisions.

“The most important factors to consider in the treatment of DCIS include the age, health and life expectancy of the patient,” Tuttle said. “The management of DCIS in a 45-year-old healthy woman is considerably different than in a 75-year-old woman with medical problems. Other factors include the family history, the presence of genetic mutations, size and grade of DCIS, and ER status. Generally, most patients can be treated with excision, with or without radiation.”


Some experts question whether patients who undergo surgery can avoid radiation.

“If low- or intermediate-grade DCIS are removed with a wide margin of normal breast tissue, without irradiation, the recurrence rates are high, and many of the recurrences are of invasive breast cancer,” Kopans said. “Our surgeons tried this and had to stop because there were too many recurrences.”

Beryl McCormick, MD, chief of the external beam radiotherapy service at Memorial Sloan Kettering Cancer Center, and colleagues conducted a prospective randomized, multi-institutional trial to evaluate the benefit of radiotherapy after surgery among patients with mammographically detected DCIS.

The analysis included 636 patients aged at least 26 years (median age, 58 years) whose DCIS measured less than 2.5 cm, with margins of at least 3 mm.

At 7 years, researchers reported a significantly lower rate of local failure rate among patients who underwent radiotherapy after surgery than among those who underwent observation (0.9% vs. 6.7%; HR = 0.11; 95% CI, 0.03-0.47). The risk for contralateral breast failure also was lower in the radiotherapy group (3.9% vs. 4.8%; HR = 1.07; 95% CI, 0.48-2.39).

Researchers reported a higher rate of toxicity in the radiotherapy arm than the observation arm (76% vs. 30%; P < .001), but incidence of grade 3 to grade 5 toxicities was 4% in each study arm. Thirty percent of patients developed grade 1 late radiotherapy toxicities, whereas 4.6% developed late grade 2 toxicities and 0.7% developed late grade 3 toxicities.

Several large prospective trials that compared breast conservation surgery with or without whole-breast radiation showed radiation reduced the risk for a local breast recurrence by at least 50%, establishing radiation as part of the standard of care, McCormick told HemOnc Today when the results were published. However, despite these results, many oncologists still consider radiation “overtreatment,” she said.

“For my practice, this is the best evidence available to discuss ‘no radiation required’ for women who present with this low-risk DCIS,” McCormick said.

Although the potential for overtreatment rightfully prompts caution, some experts suggest patients ultimately benefit from a more proactive approach.

“There is a legitimate argument to be made that a lot of women are being exposed to unnecessary and unhelpful treatment,” Smith said. “There is also a converse argument to be made that we are doing some good in identifying these DCIS cases and treating them at an earlier stage.”

Patient preferences also should be taken into account when formulating treatment plans, Smith said.

“If a patient has low-grade DCIS and her long-term risk of developing invasive disease is only 20%, but you can cut that down to 1% or 2%, that might be worthwhile for that individual patient even though she would have had an 80% chance of being fine had she [never been diagnosed],” Smith said. “These are complicated issues, and there are no simple, easy answers.”

Calling it cancer

A cancer diagnosis immediately changes a person’s life, and even the suggestion of cancer can have a similar effect.

The ensuing fear and anxiety — along with the possibilities for harmful and unnecessary treatments — have prompted many clinicians to advocate for a change in the nomenclature of DCIS.

Advocates for the name change — such as Laura Esserman, MD, MBA, director of the Carol Franc Buck Breast Care Center and co-leader of the breast oncology program at UCSF Helen Diller Family Comprehensive Cancer Center — point to concerns about the aggressive treatments that often ensue after someone is diagnosed with cancer.

“DCIS is not cancer,” Esserman told The New York Times in 2013. “It’s a risk factor. … We don’t do heart surgery when someone comes in with high cholesterol. What are we doing to these people?”


Tuttle supports reclassification, saying patients would benefit from a more clearly defined and easily understandable categorization.

“DCIS is not infiltrating ductal carcinoma,” Tuttle said. “These terms are unnecessarily confusing. Patients frequently ask, ‘Well, doctor, is it or is it not cancer?’ Patients with breast cancer and DCIS deserve accurate and easily understood information about their disease so they can make informed decisions. Perhaps a better understanding of the disease will lead to less anxiety and less aggressive treatment.”

A diagnosis of “cancer” may create a needless burden for clinicians and patients, and removing that burden could give everyone involved the ability to focus on what DCIS actually is: a potential step toward a cancer diagnosis but not cancer itself, experts said.

“I don’t think anyone would disagree that a high-grade DCIS needs to be treated seriously,” Virnig said. “There is no controversy there. But there is really a lot of discussion about whether we are turning people into patients who otherwise would never have become patients.

“We are finding DCIS that we never would have found, because we have far greater resolution now [on screening images] than we ever did with the old film,” Virnig said. “Once you find it, you can’t ‘unfind’ it.”

In the late 1980s, cervical carcinoma in situ — or stage 0 disease — became known as cervical intraepithelial neoplasia. A similar effort with DCIS, removing the “cancer” connotation, could have positive public health implications, Virnig said.

“Carcinoma means cancer,” Virnig said. “It’s a labelling issue. … Not every clinician is going to be prudent about what is appropriate treatment for the level of risk the disease actually has. It’s also expensive — both in the amount of money and the amount of time, because people’s time is worth something. … We really have to focus on differentiating between cases that we expect to be serious and those we do not.”

However, it is important to remember widespread adoption of breast cancer screening — and increased awareness of DCIS — has contributed to the decline of invasive carcinoma incidence, Kopans said.

“The incidence of invasive breast cancer has actually declined in the United States since the start of screening,” Kopans said. “This is likely due to the detection and removal of DCIS lesions due to screening going back to 1985. If the precursor lesion is removed, the rate of subsequent invasive cancers should go down, as the data show.”

Reclassification of the disease also could lead to a different type of confusion, Kopans added.

“The fundamental problem is that women and their doctors do not like to have a cancer recur,” Kopans said. “Tricking women by saying that these are just ‘high-risk lesions’ can reduce the ‘overtreatment,’ but there will be consequences from misleading some women. I have no problem renaming DCIS, although you will have to explain to women why they have developed invasive breast cancer when they initially had a ‘benign’ lesion.” – by Cameron Kelsall


American Cancer Society. Cancer Facts and Figures 2015. Available at: Accessed on June 9, 2015.

CDC. Breast cancer statistics. Available at Accessed on June 9, 2015.

Kopans DB. Oncologist. 2014;doi:10.1634/theoncologist.2013-0184.

McCormick B, et al. J Clin Oncol. 2015;doi:10.1200/JCO.2014.57.9029.

Morrow M. J Clin Oncol. 2012;doi:10.1200/JCO.2011.40.5514.

Page DL, et al. Cancer. 1982;49:751-758.

Rosenberg PS, et al. J Natl Cancer Inst. 2015;doi:10.1093/jnci/djv159.

Smith BD. J Clin Oncol. 2015;doi:10.1200/JCO.2014.59.4259.

Smith RA, et al. CA Cancer J Clin. 2003;53:141-169.

Virnig BA, et al. Am Soc Clin Oncol Educ Book. 2012;doi:10.14694/EdBook_AM.2012.32.45.

For more information:

Daniel B. Kopans, MD, can be reached at Massachusetts General Hospital, 15 Parkman St., Mailstop: Level 2, Suite 219, Boston, MA 02114; email:

Beryl McCormick, MD, can be reached at Memorial Sloan Kettering Cancer Center, 1275 York Ave., New York, NY 10065.

Philip S. Rosenberg, PhD, can be reached at NCI, Division of Cancer Epidemiology & Genetics, Biostatistics Branch, NCI Shady Grove Campus, Room 7E130, 9609 Medical Center Drive, Rockville, MD 20850; email:

Benjamin D. Smith, MD, can be reached at The University of Texas MD Anderson Cancer Center, 1515 Holcombe Blvd., Houston, TX 77030; email:

Todd M. Tuttle, MD, MS, can be reached at University of Minnesota Medical School, 420 Delaware St. SE, Mayo Mail Code 195, Minneapolis, MN 55455; email:

Beth A. Virnig, PhD, MPH, can be reached at University of Minnesota School of Public Health, 420 Delaware St. SE, Mayo Mail Code 729, Minneapolis, MN 55454; email:

Disclosure: Kopans reports inventing digital breast tomosynthesis, also known as 3-D mammography, the patent for which is held by his employer, Massachusetts General Hospital. Smith reports research funding from Varian Medical Systems. McCormick, Rosenberg, Tuttle and Virnig report no relevant financial disclosures.



Should all women with DCIS receive treatment?


All patients with DCIS should be informed that radiation therapy offers the lowest risk for recurrence in the breast.

Bruce Haffty

Bruce G. Haffty

Women undergoing breast conserving therapy for DCIS should be informed of the potential treatment options and available evidence-based data supporting those options. Although active surveillance is a potential treatment option that should be discussed with those patients who, after lumpectomy with a negative margin, have low-risk DCIS, all women with DCIS, regardless of the risk profile, should be informed that radiation therapy following breast conserving therapy offers the lowest risk for recurrence.

The fact remains that all available high-level evidence from randomized trials demonstrate a significantly lower risk for recurrence with radiation following lumpectomy compared with lumpectomy alone. Long-term updates from the Swedish DCIS trial recently published (Warnberg F, et al. J Clin Oncol. 2014;doi:10.1200/JCO.2014.56.2595.) demonstrate that radiation provides a RR reduction of 37.5% and an absolute risk reduction of 12% with 20 years of follow-up. The 15-year update of the EORTC randomized trial (Donker M, et al. J Clin Oncol. 2013;doi:10.1200/JCO.2013.49.5077.) demonstrated a 48% reduction of any local recurrence with radiation and revealed that one in three non-irradiated women developed a local relapse with observation alone. Although one might argue that these trials included a broad spectrum of patients with DCIS, and that lower-risk patients do not benefit as much, the recent publication of the RTOG 9804 (McCormick B, et al. J Clin Oncol. 2015;doi:10.1200/JCO.2014.57.9029.), a study specifically in women with low-risk DCIS, shows an advantage to radiation even in this low-risk subset. This randomized RTOG study of women with low-risk DCIS treated with lumpectomy revealed a highly statistically significant benefit to radiation therapy in reducing the risk for a local relapse from 6.7% without radiation to 0.9% with radiation, with just 7 years of follow-up. With longer follow-up, these rates are likely to further diverge. Acknowledging that the absolute risk of recurrence was relatively small in this low-risk subset, radiation therapy to the intact breast clearly provided the highest probability of local control of disease. Furthermore, a subset of patients with DCIS who do not benefit from radiation has yet to be defined.


Although I would acknowledge that discussion of all of the available treatment options should be presented to the patient, all patients should be informed that radiation therapy to the intact breast offers the lowest risk for relapse and the highest chance of local control in the breast. When all of the risks and benefits are fully explained to the patient, including the fact that radiation therapy to the breast will offer the lowest risk for recurrence in the conservatively treated breast, an informed decision can be made by the patient and her treating physicians.

Bruce G. Haffty, MD, is professor in and chairman at the department of radiation oncology at UMDNJ–Robert Wood Johnson Medical School and associate director of the Rutgers Cancer Institute of New Jersey. He can be reached at: Disclosure: Haffty reports no relevant financial disclosures.


It is likely that future management of DCIS will include consideration of surveillance for some patients with low-risk DCIS.

E. Shelley Hwang, MD, MPH

E. Shelley Hwang

Annually, approximately 65 million women undergo mammographic screening in the United States at a cost of over $13 billion. Almost one in 1,300 mammograms will detect DCIS, considered the earliest detectable form of breast cancer. Over 62,000 women in the U.S. will be diagnosed with DCIS this year alone, and almost all of these diagnoses will be made in asymptomatic individuals. Without treatment, it is estimated that only 20% to 30% of DCIS will progress to invasive cancer; unfortunately, we lack reliable markers to identify those patients who will not progress. As a result, once diagnosed, guidelines dictate that DCIS be treated with a combination of surgery, radiation and hormonal therapy — treatments similar to those recommended to patients with invasive cancer.

The term “overdiagnosis” has been used to define these conditions that look like early cancer but are not destined to cause symptoms or death during a patient’s lifetime, and this concept has been applied to many cancers detected by screening. For breast screening, overdiagnosis and overtreatment resulting from mammography have been estimated to occur in as many as one in four patients diagnosed with breast cancer, and the national health care expenditure from false-positive mammograms and breast cancer overdiagnosis has been estimated to approach $4 billion annually. There is general consensus that much of this burden derives from the treatment of DCIS.

In other disease sites, there have been efforts aimed at reducing the harms of overtreatment by employing a “watchful waiting” or “active surveillance” (AS) strategy. AS implies that no definitive therapy is undertaken upon diagnosis of low-risk conditions, but that decisions regarding whether and when to intervene with treatment are made based on disease behavior over time. For prostate cancer, AS has become widely accepted as part of routine care over the last decade, based upon the recognition that many men die with their prostate cancer, rather than from it. There are now compelling data supporting surveillance over treatment for low-risk invasive prostate cancer; studies have shown equivalent outcomes when employing a strategy to intervene only upon disease progression, as opposed to treating all patients indiscriminately upon diagnosis. This approach allows the biologic behavior of a tumor to dictate therapy, thereby reducing overtreatment by reserving treatment only for those likely to derive greatest benefit.

A similar approach has been proposed for DCIS, both in the U.S. and Europe. In the past year, two prospective randomized trials, the LORIS and LORD studies, have opened to accrual. Eligible patients are those with “low-risk” DCIS — defined as low- or intermediate-grade DCIS without invasive cancer — who are willing to consider AS for their disease. Women participating in the trials are randomly assigned to either AS or standard care, with the primary outcome being the proportion of patients progressing to invasive cancer. Such trials will help define those patients with DCIS who may be safely monitored with surveillance rather than invasive treatment, akin to the strategy for prostate cancer. With such efforts, it is likely that future management of DCIS will include consideration of surveillance for some patients with low-risk DCIS.


American Cancer Society. Cancer Facts and Figures 2014. Available at: Accessed June 18, 2015.

Bleyer A, Welch HG. N Engl J Med. 2013;doi:10.1056/NEJMc1215494.

Cooperberg MR, et al. J Clin Oncol. 2011;doi:10.1200/JCO.2011.34.9738.

Erbas B, et al. Breast Cancer Res Treat. 2006;97:135-144.

Ernster VL, et al. J Natl Cancer Inst. 2002;94:1546-1554.

Gøtzsche PC, Jørgensen KJ. Cochrane Database Syst Rev. 2013;doi:10.1002/14651858.CD001877.

Klotz L, et al. J Clin Oncol. 2010;doi:10.1200/JCO.2009.24.2180.

Ong MS, Mandl KD. Health Aff (Millwood). 2015;doi:10.1377/hlthaff.2014.1087.

Ozanne EM, et al. Breast Cancer Res Treat. 2011;doi:10.1007/s10549-011-1430-5.

Welch HG, Black WC. J Natl Cancer Inst. 2010;doi:10.1093/jnci/djq099.

Welch HG. BMJ. 2009;doi:10.1136/bmj.b1425.

Wilt TJ, et al. N Engl J Med. 2012;doi:10.1056/NEJMoa1113162.

Zahl PH, et al. BMJ. 2004;328:921-924.

E. Shelley Hwang, MD, MPH, is professor of surgery at Duke University School of Medicine. She can be reached at Disclosure: Hwang reports an advisory role with Genomic Health.

Alastair Thompson, BSc(Hons), MB ChB, MD, FRCSEd, is professor in the department of surgical oncology at The University of Texas MD Anderson Cancer Center. He can be reached at Disclosure: Thompson reports no relevant financial disclosures.