Fibroepithelial lesions (FELs) are a common histologic finding on core needle biopsy (CNB) of the breast. Fibroepithelial lesions include fibroadenoma and phyllodes tumor, which can be difficult to distinguish with an initial CNB. An institutional experience was reviewed from February 12, 2001, to January 4, 2007, to determine the safety of selective rather than routine excision of FELs and to determine the factors associated with upgrading diagnosis of FELs to phyllodes tumors without definitive phyllodes tumor diagnosis by CNB. Of 313 patients, 261 (83%) with FELs diagnosed by CNB received observation with long-term follow-up (mean, 8 years). Of the observed patients, 3 (1%) were diagnosed with phyllodes tumor on follow-up. Eighteen of 52 patients (35%) who received excision had an upgrade of diagnosis to phyllodes tumor. Sensitivity and specificity of the pathologist’s comment of concern for phyllodes tumor on a CNB demonstrating FELs without definitive phyllodes tumor diagnosis were 82% and 93%, respectively. Our policy of selective excision of FELs without definitive phyllodes tumor diagnosis resulted in safe avoidance of many surgical procedures.
The goals of clinical and imaging evaluation of the breast are to find all cancers and avoid unnecessary diagnostic surgical procedures, if possible.1 Core needle biopsy (CNB) achieves these goals.2- 5 Fibroepithelial lesions (FELs) are one of the most common lesions diagnosed by CNB.6 These include fibroadenomas and phyllodes tumors.7 Both are biphasic neoplasms containing a proliferation of epithelial and stromal components.7- 9 Imaging and histologic characteristics of these lesions can overlap.7- 11 When differentiation between them is difficult with CNB, the World Health Organization Working Group that convened in September 2011 in Lyon, France, favors a diagnosis of fibroadenoma to avoid overtreatment.7- 9
On finding an FEL, the pathologist may designate it an FEL and then add a comment of concern, such as “cannot rule out phyllodes” or “increased stromal cellularity,” if features of phyllodes are present but not definitive.11 The exact character of an FEL is important owing to differences in recommended management ranging from observation to wide surgical resection.5,12
The purposes of this study were to determine our institutional rate of upgrading FEL diagnosis to phyllodes tumor without definitive phyllodes tumor diagnosis by CNB and to determine the factors predicting upgrade. We hypothesized that a policy of selective excision of FELs diagnosed by CNB would result in safe avoidance of many surgical procedures.
Institutional review board approval was obtained from Gundersen Health System for retrospective review of our prospective database for FEL diagnosed by CNB from February 12, 2001, through January 4, 2007. The mean (median) duration of follow-up in the observed and excised groups were 8 (8.3) and 7.2 (7.9) years, respectively. Of 313 patients, 16 (5%) were lost to follow-up after 6 months. Inclusion criteria included patients with CNB that demonstrated an FEL with the descriptor “fibroadenoma” (235 [75%]) or “any additional descriptor” (78 [25%]). Three patients with definitive phyllodes tumor diagnosed by CNB were excluded. All 3 were diagnosed with phyllodes tumor after excision. Uncertainty of exact classification of an FEL diagnosis by CNB was described with a comment of concern, and lesion classification was assigned after excision.7,8 Selective excision of FELs identified by CNB was used.2,5,12 All CNBs were performed by breast specialty radiologists. The CNB specimens were reviewed by a breast specialty pathologist. A weekly imaging pathology concordance assessment conference was held.13 All patients gave informed consent regarding the options of observation or surgical excision. Observed patients received scheduled imaging.1- 5 Patients with stable or diminishing lesions returned to screening; patients with enlarging lesions were recommended for excision.
The data analysis had 2 goals. First, to determine why certain lesions were recommended for excision, we compared lesions that were excised with those that were observed. Second, to determine which lesions were at higher risk for upgrade of diagnosis to phyllodes tumor, we compared patients found to have phyllodes tumor after excision with those with fibroadenomas. The purpose of this analysis was to search for information to aid future decision making regarding selective excision. Univariate analysis included χ2, Fisher exact, and unpaired 2-tailed t tests according to normality of distribution. Wilcoxon rank sum tests were used for follow-up review case assessment. A multivariate analysis with backward elimination method was performed to assess the significance of variables in the univariate analysis. Odds ratios with 95% CIs were determined.
The outcomes of 313 patients with FELs diagnosed by CNB are detailed in the Figure. A total of 261 (83%) received observation and 52 (17%) received excision. In the observed group, phyllodes was found in 3 patients (1%): 2 at 6 months and 1 at 18 months. Of the patients who received excision, 18 (35%) had phyllodes tumors. The proportion of patients with CNB demonstrating FEL with the descriptor “fibroadenoma” without any additional descriptor was higher in patients who received observation (223 of 261 [85%]) than in those who received excision (12 of 52 [23%]; P < .001).
Linear graph with biopsy results, patient care pathway, and results of long-term follow-up. CNB indicates core needle biopsy; FEL, fibroepithelial lesion; and NCCN, National Comprehensive Cancer Network.
aEight of 261 observed patients (3%) were not followed up.
In univariate analysis, patients who received excision were younger (mean age, 38.8 vs 45.8 years; P = .001), had larger tumor size (mean, 16 vs 12 mm; P = .003), and more often had a pathologist comment of concern on CNB (mean, 67% vs 1%; P = .001). Other factors associated with excision in univariate analysis are detailed in Table 1. In multivariate analysis, comments by pathologists (odds ratio [OR], 612.99; 95% CI, 96.53-999.99), detection by providers (OR, 9.94; 95% CI, 3.05-32.41), tumor size larger than 20 mm (OR, 8.30; 95% CI, 2.46-28.0), and positive family history (OR, 3.74; 95% CI, 1.17-11.96) were associated with excision.
Factors associated with an upgrade of diagnosis from FEL to phyllodes tumor without definitive phyllodes tumor diagnosis by CNB after excision in univariate analysis are detailed in Table 2. These patients were older (mean age, 43.5 vs 36.o years; P = .03) and more often had a pathologist comment of concern on CNB (18 of 18 vs 17 of 34; P < .001). No variable was significant in multivariate analysis.
Sensitivity, specificity, and positive and negative predictive value of comment of concern on CNB were 82%, 93%, 47%, and 98.5%, respectively. After excision was discussed, 3 patients with a CNB comment of concern chose observation. None developed phyllodes tumor during follow-up.
Most FELs diagnosed by CNB are fibroadenomas. Fibroadenomas diagnosed by CNB with imaging pathologic concordance can usually be safely managed without a surgical procedure.11 On the other hand, a CNB demonstrating FELs with any histologic feature suggesting a more advanced lesion than fibroadenoma may indicate a phyllodes tumor. There are no absolute determinative histologic features to distinguish between them.7,89,11,12,14 Consequently, surgical excision is recommended for many patients, resulting in potential overtreatment.7- 9,15,16 Of our patients, 83% avoided surgery. Factors associated with excision included a pathologist comment of concern. Of patients who received excision, 35% had a phyllodes tumor. In other studies, this rate ranges from 18% to 38%.6,11,17,18 Our experience supports American College of Radiology and National Comprehensive Cancer Network recommendations for selective management of FELs identified by CNB given our missed phyllodes tumor rate of 1% during an 8-year follow-up period. Some phyllodes tumors may have been missed owing to indolent growth or loss to follow-up. Other limitations of our study include the observational study design and a few missing data points. Consistent with a recent study by Karim et al,9 we have not yet identified absolute determinative criteria for excision of FELs when definitive phyllodes tumor is not evident after CNB.
Resetkova et al11 and others13,19,20 advocate reliance on pathologist comments and concordance assessment as important factors for choice of excision for FEL and other high-risk lesions. Sensitivity and specificity regarding a pathologist comment of concern were 82% and 93%, respectively, in our experience. Patients not receiving excision of FELs required follow-up for missed lesions. Three of our 261 observed patients were later upgraded to a benign phyllodes tumor diagnosis after lesion enlargement was identified. All were identified within 18 months. Benchmarks for both the upgrade and miss rate of FELs by CNB have not yet been established. However, our results provide proof of concept that selective excision of FELs as described herein allows many patients to avoid a surgical procedure, with low risk of missed phyllodes lesions. In agreement with Karim et al,9 we encourage research to identify imaging, molecular, and morphologic factors that better discriminate FELs on CNB to increase the proportion of patients who can avoid surgical excision.
Accepted for Publication: January 6, 2014.
Corresponding Author: Jeffrey Landercasper, MD, Norma J. Vinger Center for Breast Care, Gundersen Lutheran Health System, 1900 S Ave, La Crosse, WI 54601 (firstname.lastname@example.org).
Published Online: August 27, 2014. doi:10.1001/jamasurg.2014.73.
Author Contributions: Dr Landercasper had full access to all the data in the study and takes responsibility for the integrity of the data and accuracy of the data analysis.
Study concept and design: Van Osdol, Landercasper, Andersen, Johnson.
Analysis and interpretation of data: All authors.
Drafting of the Manuscript: Van Osdol, Landercasper.
Critical revision of the manuscript for important intellectual content: All authors.
Statistical analysis: Al-Hamadani, Marcou, Vang.
Administrative, technical, and material support: Van Osdol, Landercasper, Andersen, Ellis, Gensch, De Maiffe, Marcou, Al-Hamadani, Vang.
Study supervision: Landercasper, Andersen, Johnson.
Conflict of Interest Disclosures: Dr Ellis reports being a medical consultant for Three Palm Software. No other disclosures were reported.
Funding/Support: The study was supported in part bythe Gundersen Medical Foundation/Norma J. Vinger Center for Breast Care for partial assistance with the data collection, analysis, and manuscript preparation.
Role of the Funder/Sponsor: The funding source had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication.
Previous Presentations: This study was presented in part at the Wisconsin Surgical Society 2012 Annual Meeting; November 2, 2012; Kohler, Wisconsin; and the National Comprehensive Cancer Network 18th Annual Conference; March 15, 2013; Hollywood, Florida.
Van Osdol AD, Landercasper J, Andersen JJ, Ellis RL, Gensch EM, Johnson JM, De Maiffe B, Marcou KA, Al-Hamadani M, Vang CA. Determining Whether Excision of All Fibroepithelial Lesions of the Breast Is Needed to Exclude Phyllodes TumorUpgrade Rate of Fibroepithelial Lesions of the Breast to Phyllodes Tumor. JAMA Surg. 2014;149(10):1081-1085. doi:10.1001/jamasurg.2014.73