Isolated Atypical Lobular Hyperplasia Diagnosed on Breast Biopsy: Low Upgrade Rate on Subsequent Excision With Long-Term Follow-up

After institutional review board approval, a pathology database search was performed for a study period of 191 months (January 2000–November 2016) to identify all breast core and excisional biopsies with a diagnosis of ALH from internal and referral material. Core biopsies were either ultrasound-guided or stereotactic-guided vacuum-assisted core needle biopsies. Our case selection criteria included core biopsies with a diagnosis of ALH, with no other atypical pathology present (ie, biopsies that contained ALH along with flat epithelial atypia, atypical ductal hyperplasia, LCIS, DCIS, or IC were excluded).

Atypical lobular hyperplasia was diagnosed when acini in a lobular unit had less than 50% involvement by a monotonous population of epithelial cells showing classic lobular characteristics, including discohesive cells with small, round nuclei, often containing perinuclear vacuoles.^1–5  Biopsies also containing LCIS (≥50% filling and expansion by a similar proliferation) were excluded.¹  The majority of cases were diagnosed using morphology alone; however, E-cadherin immunohistochemistry was used to confirm lobular phenotype in a minority of cases.

In order to examine the true upgrade rate of isolated ALH, we limited our cohort to cases with radiologic-pathologic concordance. Radiology reports were carefully reviewed and cases in which biopsy results were felt to misrepresent the radiologic target were excluded. Examples included suspicious calcifications that did not match those observed in the biopsy material and radiographic mass-forming lesions that were not represented in the biopsy.

Clinical characteristics including age at diagnosis, history, radiologic findings, indication for biopsy, and follow-up were obtained from patient electronic medical records. Specifically, follow-up information for disease status was obtained through the most recent radiologic studies (mammogram, magnetic resonance imaging) and clinical notes. Patients with a history of or concurrent breast carcinoma were included only if the carcinoma was in the contralateral breast or in a different quadrant in the same breast as the ALH. Excisions were deemed upgraded when a diagnosis of IC or DCIS was rendered following a biopsy diagnosis of ALH.

Descriptive statistics (mean and standard deviation for continuous data and frequency and percentage for categorical data) were presented. A simple logistic regression model was used to associate each of the risk factors to upgraded pathology on excision; the risk factors included age, location, additional pathologic findings, indication for biopsy, high-risk status (family history), and personal history. A P ≤ .05 was considered to be statistically significant. All analyses were conducted using SAS (version 9.4, SAS Institute, Cary, North Carolina).

Figure 1 illustrates the cases of isolated ALH diagnosed on core biopsy with follow-up surgical excision included in this study. Of 30 732 consecutive breast pathology cases, 87 core biopsies (0.3%) with isolated ALH were included after excluding biopsies with coexisting atypical lesions, excisional biopsies, discordant radiologic-pathologic cases, and cases that lacked surgical excision. Fourteen cases were excluded because of the patients' electing to forego subsequent surgical excision. Additionally, the majority of these patients did not continue clinical follow-up. Follow-up information (≥6 months) was available for 63 of 87 cases (72%).

Eighty-seven core biopsies from 87 patients were included. All patients were female, with ages ranging from 32 to 81 years (54.8 ± 10.1 years). Sixty-one (70.1%) had no history and 26 (29.9%) had a personal history of or synchronous breast carcinoma. The most common indication for core biopsy was mammographic calcifications (n = 71; 82%), followed by a mass or nodule (n = 6; 7%), asymmetry or distortion (n = 5; 6%), nipple discharge (n = 2; 2.5%), and magnetic resonance imaging enhancement (n = 2; 2.5%). The laterality of the targeted biopsies was essentially equivalent (right, n = 41 [47%]; left, n = 46 [53%]). Additional pathologic findings present in biopsies included fibrocystic changes (n = 73; 84%), fibroadenoma (n = 10; 11.5%), papilloma (n = 1; 1%), and no additional histologic findings (n = 3; 3.5%). Calcifications, when present (n = 71 of 87; 82%), were detected within fibrocystic changes and/or benign glands in all cases. Additionally, microcalcifications were focally present within ALH in 7 cases (7 of 71; 10%).

Twenty-six patients had a personal history of or synchronous diagnosis of breast carcinoma. Of these, 10 had a previous diagnosis of breast cancer ranging from 1 to 18 years preceding the ALH diagnosis, and the remaining 16 had a concurrent diagnosis of breast carcinoma. Overall, the breast carcinoma was located contralateral to the ALH biopsy in 19 patients, and 7 had breast carcinoma in the ipsilateral breast.

Excision findings are summarized in Table 1. The overall upgrade rate to carcinoma on surgical excision was 3.4% (3 of 87 cases). All 3 upgraded cases revealed DCIS on the excision; no cases contained IC (Figure 2, A and B). Of the 3 upgraded cases, 2 patients had a synchronous diagnosis of breast carcinoma: one with IC in the contralateral breast and the other with DCIS in a different quadrant in the ipsilateral breast. All 3 patients lacked family history of breast cancer. The indication for biopsy in all 3 cases was calcifications, which were present in the biopsies associated with columnar cell changes, fibrocystic changes, and benign glands. Two cases contained residual calcifications on postbiopsy imaging; however, these were considered concordant with the calcifications described in the biopsy (ie, residual calcifications did not drive the decision to pursue subsequent surgical excision). All calcifications were removed with the biopsy in the third case. Excision revealed 2 cases of intermediate–nuclear-grade and 1 case of high–nuclear-grade DCIS, multifocal in 2 cases and focal in 1 case, averaging 2.1 cm in microscopic size. Biopsy site changes were identified in all 3 cases surrounding areas of DCIS. Calcifications were present in DCIS in 2 cases, and no residual calcifications were identified in the third. Patient age, location, additional findings, biopsy indication, family history, and personal history were analyzed by univariate analysis to assess associations with upgraded pathology on excision. None of these variables were found to be significantly associated with an upgraded excision diagnosis. Although it was not statistically significant (P = .20), patients with a personal history of or concurrent diagnosis of breast carcinoma had a higher likelihood of having an upgraded lesion on excision (2 of 26 and 1 of 61; 7.7% versus 1.6%, respectively).

Of note, we excluded 12 cases with discordant radiology. Seven of these patients underwent immediate surgical excision, and results are as follows: IC (n = 2), DCIS (n = 1), fibroadenoma (n = 2), atypical ductal hyperplasia (n = 1), and fibrocystic changes (n = 1). In all 7 cases, a suspicious mass lesion was detected on imaging and the biopsy revealing isolated ALH was not felt to be representative of the targeted mass. Overall, 3 of 7 discordant cases (43%) with excision revealed an upgraded lesion.

Follow-up information (≥6 months) was available for 63 patients (57.8 ± 43.9 months; range, 6–183 months) and is summarized in Table 2. The majority of patients showed no evidence of further breast disease (n = 49 of 63; 77.7%) or underwent bilateral mastectomies (n = 5 of 63; 7.9%); a minority of patients developed additional atypical lesions (atypical ductal hyperplasia) (n = 1 of 58; 1.7%) or carcinoma (n = 8 of 58; 13.8%). In patients with no history of breast carcinoma and ALH diagnosed on biopsy, the majority developed IC in the contralateral breast in relation to the ALH biopsy (n = 5 of 6; 83%), with an average time interval between ALH diagnosis and breast carcinoma of 62.2 ± 42 months (range, 12–117 months). One patient developed new suspicious calcifications in a different quadrant in the ipsilateral breast 45 months following the ALH biopsy, which revealed DCIS.

Guidelines for the management of isolated ALH on breast biopsy are not well established. The lack of a conclusive recommendation can be partially explained by conflicting reports of upgrade rates in the literature. The likelihood of finding a higher-grade lesion on excision after a biopsy with ALH is reported as anywhere between 0% and 20%. These vast differences may be in part due to the selection criteria of cases studied, such as failing to exclude cases with discordant radiologic-pathologic findings. More recently, studies^12–15  with radiologic-pathologic concordance report upgrade rates of 0% to 3.5%.

In the current study, we choose to evaluate cases of pure, isolated ALH in order to better estimate the immediate upgrade rate to carcinoma of ALH on core biopsy. We found that the immediate upgrade is low in our total population, and even lower if we isolate patients without a history or concurrent diagnosis of breast carcinoma. We attempted to elucidate parameters that may lead to higher immediate upgrade rates; however, age, location, additional pathologic findings, biopsy indication, high-risk status (family history), and personal history did not pan out as significant factors. Although it was not statistically significant (P = .20), likely because of the small number of total upgrades and small overall size of our cohort, patients with a prior or concurrent history of breast carcinoma showed a higher rate of upgrade on excision compared with patients without a history of breast disease (7.7% versus 1.6%, respectively). It should be noted that because of small numbers in our study, associations with upgrades may not adequately represent true relationships.

It is clear that radiologic-pathologic correlation is crucial when evaluating the true upgrade rate of isolated ALH. In our series, discordant findings led to carcinoma upgrades in almost half of cases. Recently published studies^12–15  with adequate radiologic-pathologic correlation support that clinical and radiologic follow-up is sufficient for isolated ALH diagnosed on biopsy when no other suspicious radiologic or histologic findings are identified. Although patient cohorts differ slightly, such as in excluding patients with synchronous or history of breast cancer^13,14  or focusing on minimal (≤3 foci) ALH,¹²  the overall upgrade rate for isolated ALH in 555 excisions with adequate radiologic-pathologic correlation is 2.7% (Table 3). From these studies, the majority of cases with immediate upgrades to a worse lesion were predominantly DCIS (n = 11 of 15; 73%). The argument for conservative management over immediate excision could even stand true for patients with immediate upgrades; the majority are DCIS and will retain a good prognosis when found during later surveillance. With these studies taken into consideration, more conservative recommendations, including close clinical and radiologic surveillance and chemopreventive therapy, are advocated by experts.^16,17 

The parameters for what constitutes a high-risk lesion in the breast differ depending on medical discipline. In the radiologic literature, a high-risk lesion is defined as any histologic lesion associated with an immediate upgrade rate higher than 2%.¹⁸  In the clinical and pathologic literature, a high-risk lesion refers to whether the lesion is a marker for subsequent breast cancer risk.^2,3  Women with a diagnosis of ALH carry a relative risk of 4 to 5 of developing future breast cancer^2–7 ; however, up until recently the absolute risk has not been clear. Recently published data from a large cohort of 698 women calculate the absolute risk of developing breast cancer in patients with atypical hyperplasia, which has put the concrete lifetime risk into perspective.¹⁹  Specifically, 25 years after a diagnosis of atypical hyperplasia (atypical ductal hyperplasia and ALH) on biopsy, breast cancer (in situ or invasive) developed in 29% of women.¹⁹  Our follow-up of 63 patients with ALH on biopsy revealed 13% of women with breast carcinoma events at around 5 years post biopsy, with the majority developing in the contralateral breast.

The biologic potential of ALH as a premalignant lesion versus a generalized marker for risk is still currently debated. Large studies have shown the cancer risk in the ipsilateral breast is 2 to 3 times that of the contralateral breast, supporting ALH as a local precursor.^5,19  In other series,⁷  cancer develops in both breasts equally following a diagnosis of ALH, and it is argued that ALH should be clinically considered a marker for bilateral breast cancer risk.

Of note, nearly all patients at our institution with a diagnosis of ALH on biopsy get referred for immediate surgical excision. Therefore, we did not have a substantial cohort of patients who did not undergo excision for future breast carcinoma event comparison. Fourteen cases of isolated ALH with concordant radiologic-pathologic findings were not excised. This was largely driven by patient choice. In addition, our institution is a large referral center and a subset of patients were seen for second opinion and chose to return to their home institution for follow-up care. We do not have excision or follow-up information on these patients. Even so, of the 63 patients who had ALH excised with long-term follow-up, 13% still developed future breast carcinoma. We did not evaluate the extent of ALH on biopsy, which has been shown in some studies to be a significant factor associated with upgrade rates.^12,20  Atypical lobular hyperplasia and LCIS fall on a histologic spectrum, and whereas we limited our cohort to ALH alone, other studies that grouped LCIS and ALH under the umbrella of lobular neoplasia have shown low upgrade rates for both lesions.^17,20,21 

In conclusion, the results of this study support recommendations for more conservative management including clinical and radiologic follow-up for patients with isolated ALH on biopsy, especially in patients without a prior history of breast carcinoma. Larger studies are needed to investigate clinical and/or pathologic parameters associated with upgrade rates as well as outcomes in patients who forego excision. Careful radiologic-pathologic correlation for every case is essential when deciding patient management and key to developing generalized recommendations for ALH. Lastly, following the dogma that ALH represents a generalized marker for increased risk of bilateral breast cancer, excising the local lesion theoretically does not decrease a patient's risk of future cancer development, further supporting a more conservative approach to management.