Annals of Surgical Oncology

, Volume 16, Issue 8, pp 2264–2269

Management of Intraductal Papillomas of the Breast: An Analysis of 129 Cases and Their Outcome

Authors

  • Nasim Ahmadiyeh
    • Department of SurgeryBrigham and Women’s Hospital and the Dana Farber Cancer Institute
  • Mariana A. Stoleru
    • Department of PathologyBrigham and Women’s Hospital
  • Sughra Raza
    • Department of RadiologyBrigham and Women’s Hospital
  • Susan C. Lester
    • Department of PathologyBrigham and Women’s Hospital
    • Department of SurgeryBrigham and Women’s Hospital and the Dana Farber Cancer Institute
Breast Oncology

DOI: 10.1245/s10434-009-0534-1

Cite this article as:
Ahmadiyeh, N., Stoleru, M.A., Raza, S. et al. Ann Surg Oncol (2009) 16: 2264. doi:10.1245/s10434-009-0534-1

Abstract

Introduction

The management of intraductal papillomas of the breast has been controversial; some advocate surgical excision of all lesions despite benign pathologic features, whereas others excise only those specimens with atypia.

Methods

We conducted a retrospective review of 129 core-biopsy-proven papillomas of the breast with atypia (n = 43) and without atypia (n = 86) and determined the rate of missed carcinoma in surgically excised specimen in each group.

Results

Carcinoma was found in 22.5% of the surgically excised specimens in the atypia group (9/40) and in 3% of the surgically excised specimens in the no atypia group (1/29).

Conclusions

Our findings confirm the practice that papillomas with atypical features should be excised, and suggest that in patients with adequate follow-up, benign papillomas may be managed conservatively.

The management of breast disease diagnosed by core needle biopsy has occasionally been controversial because of the potential for understaging or a missed carcinoma due to undersampling of the lesion or adjacent tissue. Intraductal papillomas of the breast have traditionally been managed by complete surgical excision to carefully examine the entire papilloma structure and adjacent tissue for ductal carcinoma in situ (DCIS) or invasive breast carcinoma. Several case series have suggested an extremely low rate of missed coexistent atypia or malignant disease in cases of benign solitary papilloma.14 Two other reviews, including one as recent as 2008, still advocate surgical excision of all breast papilloma, regardless of histopathologic features.5,6

The recommendation for excision of benign papillomas is largely precautionary. There is consensus that papillomas with atypical pathologic features warrant surgical excision, because the rate of coexistent carcinoma in every study is quite high (22–67%); however, controversy exists regarding the optimal management of papillomas without atypia.14,7

We report our findings from a retrospective review of 129 consecutive biopsy-proven intraductal papillomas diagnosed within a 4-year period, including patients managed conservatively as well as those who underwent surgical excision. We were primarily interested in the rate of missed carcinoma in excised specimens.

Methods

Pathology records of all breast biopsies performed at Brigham and Women’s Hospital between 2004 and 2008 were searched for the term “papilloma” and yielded 223 records. Chart review of these 223 pathology records identified 129 cases in which papilloma was diagnosed on core biopsy. The remaining records were cases where the word “papilloma” appeared in the requisition/record as part of the differential diagnosis, or where papilloma was identified incidentally in a specimen excised for cancer. There were no cases of papilloma identified by excisional biopsy directly.

All breast imaging studies at our institution are reviewed by dedicated breast imagers, who perform the core biopsies with image guidance. Percutaneous core biopsies are performed using 8-, 11-, or 14-gauge biopsy devices with mammographic, ultrasound, or MRI guidance. All biopsy specimen are reviewed by pathologists with subspecialty expertise in breast pathology. Surgical excision is performed by dedicated breast surgical oncologists. In this study, we were primarily interested in the pathologic diagnosis of the surgically excised specimen for those patients who underwent surgical excision of their papillomas, as well as the clinical outcome of those who were managed conservatively after core biopsy.

Patients with diagnoses of “papilloma” on core biopsy were stratified into two groups: “papilloma with atypia” and “papilloma without atypia.” Whereas definitions of atypia vary in the literature, inclusion in the atypia group for the purposes of our study was based on pathologic diagnosis of atypical ductal hyperplasia, focal atypia, nuclear atypia, focal cribiforming (n = 40 or 93%), or other histopathologic findings (n = 3 or 7%), which the pathologists deemed were best evaluated with excision: one case with a very unusual squamous metaplasia, one case where the papilloma had no supporting stroma and minimal cribiform pattern, and one case where excessive columnar cell hyperplasia with atypia prompted the pathologist to recommend excision.5,810 There were no cases of pathologic–radiologic discordance. There were two cases of papillomas with coexistent LCIS; these cases met criteria for atypia due to the presence of other above-mentioned features and were hence included in the atypia group. Cases of coexistent DCIS or other invasive carcinoma on needle core biopsy were excluded from the study altogether. All cases were of solitary papillomas rather than diffuse papillomatosis.

Clinical variables of interest included the sentinel event prompting the core biopsy (symptoms, screening mammography), past personal history of breast disease, size of core biopsy, age of patient, and radiographic features on mammography and ultrasound.

Differences between groups were determined by two-tailed Fisher’s exact test for categorical variables and by two-tailed t-test for continuous variables, threshold for significance set at P < 0.05.

Results

A total of 129 patients had pathologic diagnosis of biopsy-proven breast papillomas during the 4-year period studied at our institution. Of these, 86 of 129 (66.7%) were categorized as papilloma without atypia, and 43 of 129 (33.3%) as papilloma with atypia (Fig. 1).
https://static-content.springer.com/image/art%3A10.1245%2Fs10434-009-0534-1/MediaObjects/10434_2009_534_Fig1_HTML.gif
Fig. 1

Schematic summarizing outcome data for 129 biopsy-proven intraductal papillomas, categorized by pathologic feature on core biopsy, and showing final pathologic diagnosis of excised specimen

Of the patients with atypia, 40 of 43 (93%) underwent surgical excision at our institution, which is the standard of care for atypia. One patient with atypia opted for radiological follow-up (now in her 47th month). Detailed chart review revealed that the decision not to excise was influenced by her primary care physician’s assessment of patient’s overall declining health status due to multiple medical comorbidities at age 83 years, and patient preference after understanding risk. The other two patients with atypia were categorized as “indeterminate,” because these two patients were referred to our tertiary care center for diagnosis and returned to their primary treatment centers for follow-up care. Therefore, we cannot accurately determine whether they underwent excision at their own treatment centers.

Of those without atypia, 29 of 86 (33.7%) underwent excision at our institution, and 42 of 86 (48.8%) were managed conservatively and with follow-up at our institution. Fifteen of 86 (17.4%) patients diagnosed with benign papilloma (without atypia) at our institution were referred from outside facilities and returned to their primary treatment centers after diagnosis. Some of these patients may have opted for excision at their outside facilities, whereas some may have opted for follow-up, but we cannot adequately assess their course, and thus categorized them as “indeterminate.”

In essence, our study was focused on the 112 patients with biopsy-proven papillomas who started and completed treatment at our tertiary-level institution. Our primary interest was the 69 patients—40 with atypia and 29 without atypia—who underwent subsequent surgical excision of their biopsy-proven papillomas to assess rate of missed or coexistent carcinoma. We also studied the outcome of the 42 patients without atypia who were managed conservatively (Fig. 1).

Rate of Malignancy in Excised Specimen

Malignancy was identified in 9 of 40 excised specimens in the atypia group (22.5%) and in 1 of 29 excised specimens in the group without atypia (3%; Fig. 1). The difference in malignancy rate between papilloma with and without atypia was significant by two-tailed Fisher’s exact test (P = 0.037). Most malignancies identified in both groups were DCIS.

Follow-Up of Those Managed Conservatively

Forty-two patients with benign papilloma on core biopsy were managed conservatively, with mean and median follow-up times of 24 months and 22 months, respectively. One patient in this group without atypia developed metastatic cancer during this follow-up period; however, she had a previous history of invasive cancer in the contralateral breast. One other patient in this group without atypia had suspicious findings on follow-up imaging, but pathologic diagnosis of lesion proved benign.

Patients with Atypia Versus Without Atypia

Table 1 compares patients with atypia to those without atypia on key demographic and clinical variables. Patients diagnosed with papilloma with atypia compared with those without atypia did not differ with regard to personal history of breast disease, whether benign or malignant, and did not differ with regard to size (gauge) of core biopsy used (mean 11.8G and 11.9G respectively, median 11G and 11G respectively).
Table 1

Patient demographic and clinical features of the 129 intraductal papillomas studied: atypia versus without atypia

Overall (n = 129)

Atypia (n = 43)

No atypia (n = 86)

P value

Age (yr)

56.91

51.95

0.023

Mean size of biopsy core (G)

11.83

11.93

0.785

Personal history of breast disease

Benign (previous benign biopsy)

3 (7%)

12 (14%)

0.353

Cancer (DCIS or invasive)

4 (9%)

12 (14%)

 

No previous breast disease

36 (84%)

62 (72%)

 

Presentation n (%)

Symptomatic

10 (24%)

30 (39%)

0.11

Asymptomatic

32 (76%)

46 (61%)

 

Mammogram (calcification vs. mass)

Calcification

20 (56%)

30 (48%)

0.532

Mass

16 (44%)

33 (52%)

 

Mean size of mass on US (mm)

10.4

9.5

0.744

For categorical variables, the absolute number/counts are shown in each cell, with percentages in parenthesis. Note that if data were not available for each variable for each patient, the absolute number/counts in each category might vary, although the percent in each category will sum to 100

Those in the atypia group were older than those in the no atypia group (mean age 57 vs. 52 years, two-tailed t test, P = 0.023). The sentinel event prompting biopsy did not differ between those with atypia and those without atypia, two-tailed Fisher’s exact test (P = 0.11). Symptoms in both groups consisted primarily of palpable mass (47.5%), bloody nipple discharge (27.5%), clear nipple discharge (10%), nipple inversion/retraction (7.5%), and breast pain (7.5%), in that order.

The primary abnormality identified on mammography, whether calcification or mass, did not differ between groups with and without atypia. Among the patients with biopsy-proven papilloma with atypia who underwent excision and had masses detected preoperatively by ultrasound, the size of the mass was significantly greater in those who were diagnosed with carcinoma versus those with benign disease (mean, 17 vs. 9 mm, two-tailed t test, P = 0.031).

Patients Without Atypia: Excised Versus not Excised

Assessment of patients diagnosed with benign papilloma (without atypia) and comparison of patients who underwent excision with those who did not undergo excision for key demographic and clinical variables is shown in Table 2. There was no discernible difference between these two groups for demographic variables of age, personal history of breast disease, or gauge of core biopsy used. Likewise, among the patients diagnosed with benign papillomas, there was no difference between patients who underwent excision and those who did not for clinical variables, such as primary abnormality identified on mammography (calcification or mass; P = 1.0), size of mass on ultrasound (P = 0.49), or clinical presentation (asymptomatic or symptomatic; P = 0.23). For example, 14 patients who presented with symptoms underwent excision, and 16 patients who presented with symptoms underwent conservative management. When looking at the distribution of specific symptoms, there was no statistically significant difference between symptom breakdown in those who underwent excision versus those who did not (Fisher’s exact test, two-tailed, P = 0.1; Table 2). Thus, among patients without atypia, no demonstrable differences could be observed on demographic or clinical parameters between those who did or did not undergo excision of their benign papillomas.
Table 2

Patient demographic and clinical variables of the 86 papillomas without atypia: excised versus not excised

No atypia (n = 86)

Excised (n = 29)

Not excised (n = 42)

P value

Age (yr)

50.72

51.53

0.82

Mean size of biopsy core (G)

12.07

11.97

0.84

Personal history of breast disease

Benign (previous benign biopsy)

6 (20.5%)

6 (11%)

0.539

Cancer (DCIS or invasive)

4 (14%)

8 (14%)

 

No previous breast disease

19 (65.5%)

43 (75%)

 

Presentation n (%)

Symptomatic

14 (52%)

16 (40%)

0.23

Asymptomatic

13 (48%)

24 (60%)

 

Symptom breakdown

Palpable mass

6

7

0.1

Bloody nipple discharge

5

2

 

Nipple discharge, nonbloody

3

1

 

Breast pain

0

3

 

Nipple retraction

0

3

 

Mammogram (calcification vs. mass)

Calcification

9 (43%)

15 (45.5%)

1

Mass

12 (57%)

18 (54.5%)

 

Mean size of mass on US (mm)

7.89

10.5

0.49

For categorical variables, the absolute number/counts are shown in each cell, with percentages in parenthesis. Note that if data were not available for each variable for each patient, the absolute number/counts in each category might vary, although the percent in each category will sum to 100

Discussion

Although proliferative diseases of the breast, most notably atypical ductal and atypical lobular hyperplasia, have been shown to be markers of an increased risk of developing breast cancer (odds ratio, 2.4–5.3), the decision to surgically excise a lesion found to be benign on core biopsy is influenced by the probability of finding missed coexistent malignant disease or by the potential for progression to malignancy.11 Many studies have shown an increased risk of coexistent carcinoma upon surgical excision of papilloma with atypia, with malignancy rates 22–67%.14,7 There is consensus that these lesions should be excised for complete pathologic diagnosis.

We also found a significantly higher malignancy rate in excised specimens of papillomas with atypia (9/40, 22.5%) versus papillomas without atypia (1/29, 3%). We found patients with atypia to be significantly older than those without atypia. Although it is likely that the mere increase in age conferred a higher baseline risk of malignancy to the patients in the atypia group, it is unlikely that age accounts for the entire difference in malignancy rates between the atypia and no atypia groups. Our findings confirm the current practice that papillomas with features of atypia on core biopsy should be surgically excised. Controversy exists surrounding the management of papillomas without associated atypia.

Recent evidence through single-institution reviews suggested that papillomas without associated atypia may be effectively managed through close clinical and radiological follow-up. Four such studies finding evidence in favor of conservative management of benign papillomas had effective sample sizes (where both preexcision and postexcision diagnosis was available for an initial diagnosis of papilloma without atypia) of 11, 13, 18, and 15.14 These studies found evidence of malignancy on excision in 0%, 0%, 3%, and 0% of cases of benign papillomas respectively.

Our study represents one of the larger series of papillomas, with and without atypia, and it too suggests that benign papillomas may be managed conservatively. Our study included 86 papillomas without atypia, 29 of which were surgically excised, and 42 of which were managed conservatively with follow-up at our institution. Of surgically excised cases without atypia, 1 of 29 or 3% was found to have malignancy at excision. This one case was DCIS within an intraductal papilloma. This patient presented with a 2.2-cm palpable mass, an unusual presentation for a papilloma, and one that would typically prompt excision by surgeons in our group.

The 42 other patients without atypia who were managed conservatively in our study were followed for a mean of 24 (median, 22) months. During this follow-up period, one patient developed metastatic breast cancer, but she had a previous history of invasive cancer. No other patient developed malignancy related to her papilloma diagnosis. Although limited by short follow-up, our study suggests that benign papillomas may not all need to be excised.

Two of the most recent studies with the largest sample sizes suggest that all papillomas, even those without atypia, be excised.7,12 The largest such study by Rizzo et al., published in 2008, consisted of 101 cases of intraductal papilloma without atypia for which surgical excision was available.7 In this study, conducted at three hospitals in Atlanta, GA, the authors found that 9 of 101 (9%) cases of papilloma without atypia were subsequently classified as DCIS at excision The primary strength of their study is the large sample size. One difference between our studies, which makes a direct comparison difficult, relates to the diagnosis of the core biopsy specimen. The paper by Rizzo et al. does not state whether the core biopsy pathologic specimens were examined by pathologists with subspecialty expertise in breast disease. In our study, every case was analyzed by dedicated breast pathologists. Because pathologic diagnosis of papillomas and classification of atypical features is challenging by core, relatively minor differences in core biopsy classification could potentially translate into different outcomes postexcision and might account for differences in rate of missed carcinoma between the two studies. Another key potential variable is size of core biopsy obtained; however, our papers do not differ in this regard. Rizzo et al. obtained 11G core biopsies; our core biopsies had a mean of 12G and median of 11G, with no difference between subcategories of atypia/no atypia or excised/not excised.

Our study, like any retrospective review, is limited by potential confounders. For this reason, it was important to see that within the no atypia group, those who underwent excision did not differ from those who did not undergo excision on key demographic and clinical variables, including age, personal history of breast disease/malignancy, size of core biopsy used, presentation of symptoms, or mammographic and ultrasound findings. Another weakness of our study is that excision of benign papillomas on core biopsy at our institution depends upon the clinician’s individual review of the case and is not standardized across our group of eight breast surgeons.

In summary, we have shown an acceptably low rate of coexistent malignancy within benign (no atypia) papillomas (3%) in concordance with four independent studies that also found extremely low rates of malignancy in benign papillomas (0-3%) but in contrast to a larger study that had a malignancy rate of 9%.14,7 All cases of malignancy identified in benign papillomas, both in our study and in the study by Rizzo et al., were cases of DCIS. Although our study could have benefitted from longer follow-up time, our findings suggest that conservative management of benign papillomas could be a viable option. A most pertinent question that remains to be answered in light of the existing controversy is whether the practice of surgical excision of all benign papillomas to identify the few missed cases of DCIS translates into any survival or quality of life advantage. If this question can be definitively answered, many women might be able to avoid unnecessary surgery. A prospective study designed to answer this question would be ideal. Our study highlights the fact that controversy regarding the ideal management of benign papillomas remains but that flexibility regarding its management should be allowed.

Copyright information

© Society of Surgical Oncology 2009