Abstract
Aim: Management of papillary breast lesions is a controversial issue, as complete excision implies surgery of numerous benign lesions. The purpose of this study was to assess concordance between percutaneous and surgical biopsy of papillomas along with factors predictive of malignancy. Patients and Methods: The study consisted of a retrospective review of papilloma cases between 2009 and 2013 at three breast cancer centers. All cases of papilloma histologically diagnosed by percutaneous biopsy and confirmed by surgical specimen were included. The biopsy results were compared with final surgical pathology. Lesion size and clinical and radiological features were recorded. Results: A total of 259 cases were included (188 simple and 71 complex papillomas). Concordance between histology after percutaneous and surgical biopsy was lower for complex papillomas, regardless of type, than for simple papillomas (p<0.001). The risk of having a complex papilloma was shown to be significantly higher in postmenopausal patients (p=0.023), and was 20 times higher if the percutaneous biopsy was malignant as opposed to benign (p<2.10). However, the false-negatives for percutaneous biopsy in complex papilloma cases were mainly related to atypical and in situ lesions. Conclusion: Percutaneous biopsy does not appear adequate for identifying papillomas requiring surgical excision due to the risk of underestimation of cancerous lesions. However, certain factors predictive of malignancy may assist with surgical management, such as age and menopausal status, lesions peripheral to the nipple, and atypia on percutaneous biopsy.
Papilloma of the breast is a benign tumor arising from a vascular feeding pedicle attached to the wall of the lactiferous ducts situated between the retroareolar region and terminal duct lobular units (1-3). The mean incidence of papilloma of the breast is estimated at 3% (3). A distinction should be made between papilloma and multiple papillomatosis, which is a proliferative papillary lesion. Histologically, the latter includes papillomas, cysts, sclerosing adenosis, apocrine metaplasia, dense fibrous stroma and papillary ductal hyperplasia with varying atypia.
Management of breast papilloma is a controversial issue. Atypical or malignant lesions can be present, sometimes arising adjacent to the papilloma, and missed on percutaneous biopsy in 0 to 29% of published cases, thereby justifying the rule of complete excision of all diagnosed papillomas (2, 4, 5). However, recent publications suggest mere radiological follow-up without surgery for papillomas with no coexisting lesions diagnosed by percutaneous biopsy if imaging results are concordant (4, 6-9).
In our retrospective multicenter study, we sought to assess concordance between percutaneous biopsy and surgical biopsy of papillomas over a 5-year period, and the clinical and radiological factors predictive of malignancy.
Patients and Methods
Following a retrospective review of papilloma cases referred to the centers of Rennes, Poitiers, Tours and Orléans, France, between January 1, 2009 and December 31, 2013, we were able to include all cases of papilloma of the breast histologically diagnosed on initial percutaneous biopsy and histologically confirmed by surgical specimen. The study population included patients who had consulted for an incidental finding on a screening mammogram or onset of clinical symptoms. Cases in which the pathology of the initial biopsy or surgical specimen was unknown were excluded. For each case, the clinical, imaging and histological data were collected from the medical files. The clinical data were collected during the initial examination. All interview data (age, personal history of breast cancer, high-risk fibrocystic breast disease or papilloma, family history of breast cancer, hormonal status) and data from the physical examination (mass syndrome, nipple discharge) were recorded. The imaging findings from the initial assessment included the mammographic appearance, location and American College of Radiology (ACR) classification, as well as the ultrasound lesion size. Lesions that were less than 3 cm from the nipple were classed as ‘central’ and those further than 3 cm away as ‘peripheral’.
The initial histological diagnosis was based on a percutaneous biopsy sample. For this study, the device type, biopsy needle gauge and imaging technique used to guide the procedure were recorded. By convention, papillomas with no concurrent lesions were referred to as ‘simple papillomas’. When a lesion coexisted with the papilloma (invasive or in situ cancer, or presence of atypia), the term ‘complex papilloma’ (10) was used.
The statistical tests used for the univariate analysis were Student's or Welch's t-test for quantitative variables, and the Chi-square test for qualitative variables. For the multivariate analysis, the binary logistic regression method was used. Correlations were considered to be significant when p<0.05.
Results
Over the study period, we collected data for 359 cases of papilloma. One hundred of these cases were excluded, 73 due to lack of surgical biopsy information and 27 due to lack of diagnostic percutaneous biopsy information, including one case in a male patient. A final total of 259 patients were included, 135 in Rennes, 69 in Tours and Orléans, and 55 in Poitiers. Following percutaneous biopsy, 189 simple papillomas (73%) and 70 complex papillomas (27%) were recorded with the following distribution: 49 (19%) with atypia, 13 (5%) with an in situ component and eight (3%) with invasive cancer. After surgical biopsy, 188 simple papillomas and 71 complex papillomas were identified. The distribution of complex papillomas was as follows: 29 (11%) with atypia, 27 (10%) with carcinoma in situ, and 15 (6%) with an invasive component (Figure 1).
The first analysis compared the population whose pathological findings were concordant (results obtained by percutaneous biopsy comparable to those of the subsequent surgical biopsy), with those whose findings showed discordance between the percutaneous biopsy and surgical biopsy. On univariate analysis, the populations were comparable in terms of overall clinical and imaging features, and personal and family history, with the exception of age which was significantly younger for the concordant cases compared with the discordant cases (54 years versus 58 years, respectively, p=0.028). The type of percutaneous biopsy, microbiopsy (core-needle 14 to 18G biopsy) or macrobiopsy (vacuum-assisted biopsy 8 to 11G), had no impact on concordance between the percutaneous and surgical biopsies. For the simple papilloma cases, concordance between percutaneous and surgical biopsy was high (84%), and significantly more frequent than for the discordant cases (35%, p<10−15). Conversely, for the complex papilloma cases, concordance between percutaneous and surgical biopsy was poor: discordance in 65% of cases compared with 16% concordance (p<10−15). This high rate of discordance was mainly due to atypical lesions (55% discordance versus 8% concordance, p<10−13).
Logistic regression showed a three-fold greater risk of discordance between biopsy results for complex papilloma compared with simple papilloma [odds ratio (OR)=3.53, 95% confidence interval (CI)=1.77-7.07, p=0.00033].
Regarding the cases of discordance between percutaneous and surgical biopsy, false-negatives (i.e. benign percutaneous biopsy but malignant surgical biopsy) accounted for 21 cases: 13 cases (62%) for atypical lesions, seven cases (33%) for in situ lesions and one case (5%) for invasive lesions. False-positives (positive percutaneous biopsy but benign surgical biopsy) accounted for 20 cases: 17 cases (9%) for atypical lesions, three cases (2%) for in situ lesions and no cases for invasive lesions. There was no significant clinical or radiological feature enabling us to identify false-positive or false-negative cases in the overall study population.
We then compared the features of the complex papillomas (n×71) and simple papillomas (n×188). The mean patient age was significantly higher for those with complex papilloma (59.3±12.2 years) than for those with simple papilloma (53.5±11.8 years) (p=0.0005), with more postmenopausal patients [52 (74%) versus 108 (59%), p=0.03]. Nipple discharge was less common in patients with complex papilloma than in those with simple papilloma [10 (14%) versus eight (31%), p=0.0042]. This was also true in cases of bloody discharge: two patients (3%) with complex papilloma had a bloody discharge versus 24 (13%) with simple papilloma. Clinical mass syndrome was no more common in complex papilloma than simple papilloma cases (p=0.18). There was no difference in radiological criteria between the complex and simple papilloma populations i.e. in high ACR classification (rating of ACR 4 or ACR 5), mammographic appearance (mass syndrome, architectural distortion, microcalcifications), and ultrasound size. On the other hand, the findings revealed a statistical difference in lesion location: 34% of the complex papillomas were peripheral versus 18% of simple papillomas, p=0.04) (Table I).
Using logistic regression, two parameters were found to increase the risk of complex papilloma: menopause (OR=2.03, 95% CI=1.12-3.82, p=0.023), and malignant percutaneous biopsy (OR=20, 95% CI=10.2-40.8, p<2 10−16).
Study flowchart.
Discussion
Our study confirms the need for surgical biopsy in the case of a percutaneous biopsy revealing a complex papilloma, particularly in postmenopausal women. The reliability of percutaneous biopsy (Table II) is poor for a diagnostic procedure, producing 11% of false-negatives. However, the false-negatives were found to be mainly due to atypical and in situ lesions, which are slow growing and have a favorable prognosis. A surgical biopsy must be undertaken if the percutaneous biopsy reveals a complex papilloma with an invasive component, and surveillance could be considered as a minimum requirement in other cases, particularly in young patients, as well as in those with peripheral lesions.
Several earlier studies showed that lesions concurrent with papillomas were mainly early-stage (atypical or in situ) lesions as observed in our study. This has two contradictory implications: (i) As our study showed, there is a high false-negative rate for percutaneous biopsies (41 cases of discordance between percutaneous and surgical biopsy findings, including 21 false-negatives in our series). A surgical biopsy therefore tends to be offered systematically even if the percutaneous biopsy results are reassuring. (ii) As observed in our series and literature reports (11-14), the majority of false-negatives concern atypical lesions. Histological diagnosis of these lesions on microbiopsy is difficult and their potential for progression is uncertain. Additionally, in the event of progression to cancer, invasive malignant lesions have a good prognosis. Consequently, at the very least, if surgery is not performed for a papilloma diagnosed as benign at percutaneous biopsy, we feel that clinical surveillance with ultrasound and mammography should be offered (7).
Items considered in the study patients.
Statistical performance of percutaneous biopsy.
Management of papilloma of the breast remains a controversial issue. Proponents of systematic surgical excision of papillomas (5, 12, 13, 15, 16) point to the risk of underestimation of malignant lesions adjacent to a papilloma which are not sampled by the percutaneous biopsy (17, 18), and also to the risk that atypical lesions concurrent with papilloma will transition to cancer (19). This attitude imposes surgery that is of no benefit for simple papillomas, which represent the majority as seen in our series (73% of cases referred to our centers). Identification of factors predictive of underestimation could help to sort patients with papilloma into suitable surgical groups.
Advanced age is a recognized factor in numerous studies, with its corollary of menopausal status, which we noted in our study (20). Cancer observed in papilloma generally has a good histological prognosis (little proliferation and well-differentiated with hormonal receptivity), being the result of a natural process that began with dysplasia (14). The more peripheral location of complex papillomas (more than 30 mm from the areola) observed in our study was noted in several earlier studies (14, 21). This may be explained by the fact that multiple papillomatosis, with its high risk of coexisting cancer and progression to cancer, is of peripheral rather than central location. Lesion size was also mentioned as a risk factor, with a threshold set at 10 mm (14). This was not a finding in our series, where the majority of patients treated were over 50 years of age (from routine screening programs) and half had no clinical symptoms (mass syndrome or discharge) whatsoever and presented with a small papilloma (mean ultrasound size 11 mm). Conversely, other studies noted that the presence of a clinically detectable mass was not a risk factor for malignancy (14, 21). Furthermore, routine screening in patients aged over 50 years in France is a possible explanation for the fact that the complex papillomas in our study were less often responsible for nipple discharge (including bloody discharge) than benign papillomas, which runs counter to conventional symptomatology (4). The ACR classification is not unanimously accepted in the literature for prediction of malignancy when faced with a papilloma. Some studies reported a high percentage of malignancy in ACR 4 and 5 grade papillomas (14, 15, 22), which is not the case in other studies (23), nor in our own series. In our study, however, the small mean size of lesions may explain the difficulty with grading the images observed according to the ACR criteria. The percutaneous biopsy technique and, above all, the size of the specimens is a recognized factor for diagnostic accuracy in the literature (12, 17). This logical factor was not found in our study, where the percutaneous biopsy technique did not affect the quality of the pathological diagnosis, probably due to the limited number of macrobiopsies taken (ultrasound-guided biopsies being more common in the case of an ultrasound target as is frequently the case with papillomas). On the other hand, the literature suggests that a study of cytokines p63/CK5/6 combined with a conventional histological analysis can indicate risk for future malignancy (24), although this was not considered in our study.
Several aspects of our study warrant discussion. Limitations include its retrospective nature and the lack of a histology slide review of the percutaneous biopsies in particular (although discordant cases were reviewed by expert pathologists), a potential ‘center effect’ with prior patient selection, notably regarding the risk of breast cancer, and the small percentage of invasive malignant lesions coexisting with papilloma.
In conclusion, percutaneous biopsy alone should not be used to identify papillomas requiring surgical treatment due to the risk of underestimation of cancerous lesions. However, some factors predictive of underestimation can guide surgical management. These include age, menopausal status, lesions peripheral to the nipple and atypia on percutaneous biopsy. Although the recommendation for a systematic surgical biopsy of any type of papilloma still appears valid, several prognostic factors based on larger series should be considered, such as a nomogram able to assist with surgical decision making (25). Furthermore, if surgery is not performed, surveillance is essential, although the procedure and various elements need to be defined (type of examination, frequency and length of follow-up according to patient age).
Acknowledgements
Acknowledgements are due to Mrs Tracey Wescott and Odile Audrain for their assistance in writing this article.
Footnotes
↵* These Authors contributed equally to this study.
Conflicts of Interest
The Authors declare no conflict of interest in relation to the subject of this article.
- Received August 25, 2015.
- Revision received September 28, 2015.
- Accepted October 7, 2015.
- Copyright© 2015 International Institute of Anticancer Research (Dr. John G. Delinassios), All rights reserved
References
In this issue
Jump to section
Related Articles
Cited By...
- No citing articles found.