Surgical De-Escalation for Benign High-Risk Breast Lesions
Surgical de-escalation is part of a larger movement of de-escalation of multidisciplinary breast cancer treatment. The challenge is to balance the outcomes.
This article first appeared in the Cleveland Clinic Journal of Medicine.
Cleveland Clinic is a non-profit academic medical center. Advertising on our site helps support our mission. We do not endorse non-Cleveland Clinic products or services Policy
One focus of the article by Vegunta and colleagues1 in the July 2023 issue of the Cleveland Clinic Journal of Medicine is whether benign proliferative lesions such as atypical hyperplasia diagnosed on core needle biopsy (CNB) require surgical excision. The estimated upgrade rate—that is, finding breast cancer at surgical excision—is variable, and consensus recommendations for an acceptable threshold for excision are emerging.2 As the sensitivity of breast imaging has improved, more benign lesions are being found,3–6 and rates of upgrade have been decreasing.
Surgical de-escalation is part of a larger movement of de-escalation of multidisciplinary breast cancer treatment. The challenge is to balance oncologic outcomes with surgical morbidity and quality of life. In this case, the de-escalation may be preceding consensus on upgrade thresholds, definitions, standardized clinical workflow, agreement on follow-up, and incorporation of patient preference.
Imaging-guided CNB to assess abnormalities detected on breast imaging has been the standard of care for decades. From 1 to 2 million benign and high-risk CNBs are performed annually in the United States. 7,8 Clear, accepted clinical guidelines are followed for the management of malignant lesions, but management of high-risk lesions differs among institutions. Further, the patient’s level of risk and risk tolerance needs to be considered. The question is whether there are currently enough data so that a “recommendation against excision” can be made. One final concern is that surgical de-escalation may actually contribute to disparities.
The history of the surgical management of breast cancer is a continuum of de-escalation. The early Halsted radical mastectomy, developed in 1894 and used for decades, was a disfiguring surgery removing the breast, all axillary lymph nodes, and the chest wall musculature. Later in the 20th century, it was replaced by the simple mastectomy (sparing the chest wall musculature and axillary lymph nodes) after results of a national trial showed equivalent survival.9 Toward the end of the 20th century, studies showed breast conservation (partial mastectomy with clear margins) and radiation to be noninferior to mastectomy for early-stage disease.10-12
The surgical management of the axilla was the next area of de-escalation, with trials showing equivalent outcomes with sentinel lymph node biopsy and axillary dissection in early-stage breast cancer.13-16 Simultaneously, de-escalation of radiation therapy for breast conservation was investigated. Shortened courses of radiation (three weeks compared with five weeks), partial breast irradiation, intraoperative radiation therapy, and the option of excluding radiation therapy in select patients (over age 70) have been explored and are finding their places.12,17-20
Future areas of de-escalation of surgery include active surveillance for ductal carcinoma in situ.21-23 Cryoablation is also being investigated.24 Large randomized controlled trials documenting the safety and efficacy of these approaches have preceded and should precede clinical adoption.21-23
Women age 60 and older represent 59% of invasive breast cancer cases, and more than 30% occur in women age 70 and older.25 Many trials involving de-escalation have resulted in age 70 as a threshold for alternative treatment approaches that are appropriate for most but not all older women. The US Social Security Administration provides an online life-expectancy calculator for citizens to estimate their remaining life span and plan for retirement. An average 70-year-old female has an estimated life expectancy of 17.6 years to an estimated life span of 87.6 years. An average octogenarian has an estimated life expectancy of 10.2 years to 90.2 years, and an average 90-year-old has an estimated life expectancy of 5.1 years to 95.1 years. A healthy 70-year-old may still have a significant risk of recurrence. Both disease-free survival and overall survival should be part of the shared decision-making discussion, particularly in healthy older women.
As one example of de-escalation, the Society of Surgical Oncology Choosing Wisely campaign of 2016, an initiative of the American Board of Internal Medicine Foundation, encouraged the advancement of a national dialogue on avoiding “…sentinel node biopsy in clinically node-negative women ≥ 70 years of age with early stage hormone receptor positive, HER2 negative invasive breast cancer.”27 Patients, however, are hesitant to de-escalate cancer therapy.28 A survey of newly diagnosed patients showed that 53% accepted aggressive treatments with significant side effects for a three-month benefit in survival.29 It has been suggested that an upgrade of 3% or less could be a reasonable threshold for offering surveillance in place of surgery,30 although it remains to be seen whether women with benign atypical lesions will accept this threshold for risk tolerance. Thresholds for excision based on limited evidence are concerning, and anticipated regret is a real and powerful driver of patient choice.31
The perception among patients and providers, however, may be that immediate surgical excision avoids underdiagnosis and undertreatment of malignancy. Well-defined, evidence-based criteria for the selection of patients for nonoperative management would help address these concerns.
Active surveillance could first be offered to patients who would have been offered nonoperative management in prospective multi-institutional trials. Two small such trials suggest that an upgrade rate of 3% or lower could be a reasonable threshold for offering surveillance vs. surgery.32-33 The first is a prospective registry of 77 patients with pure lobular neoplasia (atypical lobular hyperplasia or lobular carcinoma in situ) who had an upgrade rate of 1% to 3%. The study also includes a literature summary of upgrade rates ranging from 0% to 27% in small retrospective single-institution studies, thereby demonstrating the need for trials with prospective data.32
The second registry involved 116 patients with papillomas without atypia, 66% of whom presented with mammographic mass or distortion, showing a 1.7% upgrade rate (2/116).33 The 3% threshold is similar to the upgrade rate of less than 2% for Breast Imaging Reporting and Data System Category 3 lesions recommending short-term follow-up with repeat imaging at 6 months as an alternative to biopsy, as the lesion is felt to have a less than 2% chance of being malignant.34 Individual institutions embarking on processes for determining radiologic-pathologic concordance must agree on patient selection, imaging findings, sampling issues, and expected follow-up. It is also important to remember that the recommendation for observation does not preclude a later recommendation for surgical excision, should findings change.35 The stated concerns of proponents of surgical de-escalation involving benign high-risk lesions are those of overdiagnosis and overtreatment. Overdiagnosis refers to biologically indolent cancers that may not go on to cause the individual harm,36 as evidenced by the increased rates of ductal carcinoma in situ detection resulting from improved mammographic screening without resultant increases in invasive breast cancer or breast cancer mortality.37 It is important to note that this could also be viewed as early diagnosis, but may lead to falsely improved survival statistics given potential lead-time bias. The US Preventive Services Task Force in 2016 set forth de-escalating screening guidelines that women begin mammograms at the age of 50 and continue every other year until age 7438 because of concerns regarding overdiagnosis, despite evidence supporting similar mortality reduction with screening mammography in women ages 40 to 49.39 In May 2023, after recognizing that mammograms starting at age 40 and modeled every other year to save (conservatively) 19% more lives, the US Preventive Services Task Force changed its recommendations to starting at age 40, yet they still recommend screening every other year.40 The National Comprehensive Cancer Network41 and the American College of Radiology42 continue to recommend annual mammograms beginning at age 40.
Accepted guidelines exist for margin width, adjuvant radiation, and sentinel lymph node biopsy in the cancer setting. However, guidelines differ for surgery vs. observation for benign high-risk lesions.43-46 Benign lesions on CNB for which surgical excision was historically recommended include atypical hyperplasia (both ductal and lobular), lobular carcinoma in situ, radial scars, and papillary lesions.41 Though the 2016 American Society of Breast Surgeons proposed guidelines47 suggested observation as an option for all but atypical ductal hyperplasia, pleomorphic lobular carcinoma in situ, and papillomas with atypia, the guidelines were not widely adopted. The more conservative National Comprehensive Cancer Network guidelines now recommend that atypical lobular hyperplasia/lobular carcinoma in situ, if radiologically and pathologically concordant and adequately sampled, can be observed for a period of one year in select patients (undefined) or excised, at the surgeon’s discretion.41 Screening magnetic resonance imaging (MRI) is not mentioned despite recommendations of the American College of Radiology to offer MRI screening to such patients.42
The concept of radiologic-pathologic concordance is difficult to define. Atypical lobular hyperplasia and lobular carcinoma in situ are felt to be incidental findings on performed CNBs as a result of imaging abnormalities. It is unclear how incidental findings can explain imaging abnormalities. There is also no consensus on adequate sampling (core needle size, number of passes, and degree of lesion removal), whether there is pathologic reporting regarding the extent of the abnormality, and whether the mode of detection is relevant. Some authors recommend observation for high-risk lesions in cases involving microcalcifications on a screening mammogram in an asymptomatic woman of average risk. Other authors suggest biopsy of mass lesions and architectural distortion on mammograms. Studies have dissimilar inclusion criteria, and rates of upgrade vary widely.30 Some studies include masses or non-mass-like enhancement on breast MRI (in high-risk patients by definition). More recent studies have not included cases with these latter findings as true upgrades, partially explaining the trend toward lower upgrade rates in recent literature.
Further, subsequent high-risk screening recommendations are inconsistent, and the uptake of preventive medication is classically poor.48-50 Many patients are noncompliant with follow-up recommendations (even for Breast Imaging Reporting and Data System-3 imaging studies with short-interval follow-up recommended).51 Few small prospective studies of observation with limited follow-up have been published and do not seem to be generalizable to different practice settings.51-55 For instance, Middleton et al52published a series of 104 patients with pure lobular neoplasia followed for a median of 3.4 years: 5 patients were subsequently diagnosed with breast cancer (3 of 5 at an unrelated site). Laws et al53 noted that in their high-risk clinic where MRI screening is not routinely recommended and following multidisciplinary discussion of all benign high-risk lesions, atypical lobular hyperplasia and classic lobular carcinoma in situ have been safely managed thus far without surgical excision based on 80 patients with pure lobular neoplasia and median follow-up of 27 months.53
Another study examined 478 patients with 483 atypical ductal hyperplasia lesions; 309 were observed and 174 underwent excision.54 With a median follow-up of 5.2 years, two cancers were identified at the index site in the surgery group (1.5%) and three in those observed (1.2%).54 A prospective study successfully triaged patients to surgery vs. observation following the establishment of predefined firm guidelines and performance of rigorous radiologic-pathologic correlation.55
Finally, it must be considered that women of color and low socioeconomic means do not receive optimal care. It has been demonstrated that Black women are more likely to be screened at nonaccredited facilities, without current equipment (including digital breast tomosynthesis, much less dedicated breast MRI), and with fewer resources for follow-up.56-,57 Disparities in uptake to MRI have been demonstrated according to educational level.58 Disparities in cancer treatment that have been demonstrated include lower rates of genetic testing in high-risk individuals,59 delays in diagnosis,60 and less appropriate surgery, radiation, and chemotherapy.61,62 Adherence to endocrine therapy in the cancer setting is suboptimal,63-66 perhaps in part owing to insurance coverage that also impacts MRI screening and uptake of and adherence to risk-reducing medication in following patients with benign high-risk lesions. Owing to these stated concerns, careful observation of benign high-risk lesions in women of low socioeconomic status may be destined for failure due to insurmountable social barriers.
In summary, the potential for upgrade to malignancy at surgical biopsy remains the principal reason for excision of benign high-risk lesions detected on CNB. In the authors’ opinion, the recommendation for observation of such lesions may not be ready for widespread implementation. Appropriate surgical de-escalation requires data demonstrating lack of utility of a given intervention combined with an informed shared decision-making discussion with the patient and standardized processes in place to assure quality.
Presently, upgrade rates in the literature are variable and have an unacceptably broad range, criteria for patient selection vary, consensus statements are vague, institutions with multidisciplinary discussions of radiologic-pathologic concordance are the exception, and patients not referred for surgical consultation (particular in lower socioeconomic groups) may have reduced access to and lowered rates of adherence to appropriate imaging and preventive strategies. While many institutions have adopted observation for benign atypical lesions, long-term data on oncologic safety are lacking.
Overdiagnosis and overtreatment are of concern and add to healthcare costs and patient morbidity, but de-escalation in this setting will take time for agreement and standardization, and concern remains regarding appropriate follow-up, particularly in vulnerable populations. Offering surveillance for high-risk lesions identified by CNB is a practice change that may be premature for many institutions.