| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Original Research |
1 Department of Diagnostic Radiology and Organ Imaging, Rm. 27029, Prince of
Wales Hospital, Ngan Shing St., Shatin, Hong Kong.
2 Department of Radiology, Alice Ho Miu Ling Nethersole Hospital, Hong
Kong.
Received December 17, 2004;
accepted after revision March 22, 2005.
Address correspondence to W. W. M. Lam
(wynnie{at}cuhk.edu.hk).
Abstract
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
MATERIALS AND METHODS. Clinical records and mammographic and sonographic findings of 40 women with papillary lesions in the breast were retrieved. The imaging features and cytologic findings were correlated with histologic findings.
RESULTS. Fifty-six papillary lesions in 40 patients underwent either mastectomy, segmental duct resection, or excision biopsy. There were three papillary carcinomas, 13 papillaryal lesions with carcinoma in situ, one atypical papilloma, four sclerosed papillomata, and 35 papillomata. Of these lesions, 37.5% (21/56) and 82.1% (46/56) could be detected on mammography and sonography, respectively. Galactography and dilated ducts helped to suggest the papillary nature of the lesions. However, mammography and sonography were not able to predict malignancy (sensitivity, 69% and 56%, respectively; specificity, 25% and 90%; positive predictive value [PPV], 60% and 75%; and negative predictive value [NPV], 33% and 90%). Combined interpretation of mammography and sonography gave a sensitivity of 61%, specificity of 33%, PPV of 85%, and NPV of 13%. Fine-needle aspiration gave a sensitivity of 44%, specificity of 68%, PPV of 31%, and NPV of 79%, whereas core biopsy gave a sensitivity of 82%, specificity of 100%, PPV of 100%, and NPV of 83% in the diagnosis of malignancy.
CONCLUSION. Radiologic features are not sufficiently sensitive or specific to differentiate benign from malignant papillary lesions. Fine-needle aspiration and core biopsy have pitfalls, and the need for surgical excision of all papillary lesions should be revisited.
Keywords: breast • cancer • mammography • papillary lesions • sonography
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Mammographic Interpretation
Images were obtained in two standard planes (mediolateral oblique and
craniocaudal). Dedicated film-screen equipment (Senographe DMR, GE Healthcare)
was used. The radiologists were informed of the study design but were blinded
to the clinical findings, full histologic diagnosis, and clinical outcome. The
mammographic findings were reported according to the American College of
Radiology (ACR) Breast Imaging Reporting and Data System (BI-RADS)
[3]. The presence of a mass,
shape of the mass (round, oval, lobular, or irregular), calcification (benign,
intermediate, or higher probability of malignancy), and the margin
characteristics (circumscribed, microlobulated, obscured, indistinct, or
spiculated) were noted.
Any lobular or irregular mass, any presence of calcification of intermediate or high probability of malignancy, and any mass showing microlobulation or an obscured, indistinct, or spiculated margin was considered suspicious of malignancy. When there was a single ductal opening with bloody or serous discharge on the day of examination, galactography was performed. Galactography was performed by cannulation of the ductal opening using a 30-gauge Jabczenski cannula (Cook). Nonionic iodinated contrast material (iopamirol 300) was injected until the patient felt discomfort or pain (0.5-2 mL). Two mammographic views (craniocaudal and mediolateral oblique) were then obtained after the injection of the contrast material.
Sonographic Interpretation
All sonographic examinations were performed with a 10-MHz linear array
transducer in the VST Master's Series (Logic 700, GE Healthcare). All
examinations were performed by one of the six attending radiologists, each
having had more than 2 years' experience in the interpretation of breast
images. All static and color Doppler images of the lesions and axillary lymph
nodes identified were saved and filmed.
All hard-copy films were reviewed by two radiologists at least 2 days after mammography interpretation. Sonographic features were analyzed according to the ACR BI-RADS US lexicon classification [3]. The presence of a mass, shape of the mass (oval, round, irregular), orientation of the lesion (parallel or not parallel to the skin line), margin (circumscribed or not circumscribed, which was further divided into indistinct, angular, spiculated, or microlobulated), lesion boundary (abrupt interface or echogenic halo), echo pattern (anechoic, hyperechoic, complex, hypoechoic, or isoechoic), posterior acoustic features (no posterior acoustic features, shadowing, enhancement, or a combination), surrounding tissue changes (duct changes, Cooper's ligament changes, edema, architectural distortion, skin thickening, skin retraction, or irregularity), calcification (macrocalcifications, microcalcification within or outside the mass), and vascularity (not present, present in lesion, present immediately adjacent to lesion, or diffusely increased in surrounding tissue) were all recorded. The presence of any of the following was considered suspicious of malignancy: an irregular mass, any lesion not parallel to the skin line, any noncircumscribed margin, a complex echo pattern, posterior acoustic shadowing or a combined pattern, Cooper's ligament changes, skin thickening, edema, architectural distortion, or skin retraction or irregularity.
The presence and shape (oval or round) of axillary lymph nodes and suspicious features (loss of echogenic hilum, presence of central necrosis) were also recorded.
Fine-Needle Aspiration or Core Biopsy
Aspirates and tissue cores obtained were retrieved. All specimens were
reviewed by two experienced pathologists, each having more than 5 years'
experience in breast pathology. The findings of cytologic specimens were
classified into four categories: insufficient to make a diagnosis, benign,
intermediate with atypia, and malignant. Lesions with atypia were considered
suspicious and the lesion was excised for further evaluation at our center.
Lesions with atypia were categorized as malignant for the following
analysis.
All fine-needle aspirations and core biopsies were performed under sonographic guidance. A 21-gauge hypodermic needle connected to a 20-mL syringe was used for fine-needle aspiration. Suction was maintained until aspirate was noted in the needle hub. For all core biopsies, a16-gauge biopsy needle was used. At least three good cores of tissue were obtained for each lesion.
Surgical Excision or Mastectomy
Decisions as to surgical excision or mastectomy were made by the attending
surgeon. Histology from surgical excision of the mass or mastectomy served as
the gold standard.
Statistical Analysis
Imaging features and cytologic findings were correlated with histologic
findings. Imaging features and cytologic findings were also correlated with
clinical outcomes. All statistical analysis was performed using SPSS version
11.0 software installed on a PC. All results with a p value of less
than 0.05 were considered significant.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Ten mastectomies were performed in these 40 patients because of breast malignancy, although some of these patients had coexisting benign and malignant breast lesions. A total of three papillary carcinomas, 13 papillary lesions associated with carcinoma in situ, and five papillomata were yielded. None of these patients had histologic metastasis to axillary lymph nodes. Thirty-seven excisional biopsies or segmental duct resections were performed; two excisional biopsies were performed before mastectomy. The remaining 35 excisional biopsies yielded a total of one atypical papilloma, four sclerosed papillomata, and 30 papillomata. A total of 16 malignant and 40 benign papillary lesions were analyzed.
Correlation of Fine-Needle Aspirates and Core Biopsy with Mastectomy or Surgical Excision
Of the 56 lesions, 41 lesions had sonographically guided fine-needle
aspiration before excisional biopsy or mastectomy. Four aspirates had
insufficient material for diagnosis and were excluded from the following
analysis. Of these 37 lesions, nine lesions were malignant and 28 were benign.
When the presence of cellular atypia was considered suspicious for malignancy,
fine-needle aspiration gave a sensitivity of 44% (95% confidence interval
[CI], 12-77%), specificity of 68% (51-85%), positive predictive value of 31%
(6-56%), and negative predictive value of 79% (63-95%).
Core biopsy was performed in a total of 21 lesions, of which 11 were malignant and 10 were benign. Three cores of tissue were obtained for each of 15 lesions, of which 10 lesions were malignant and five were benign. Four cores of tissues were obtained for each of the remaining one malignant and five benign lesions. Papilloma with foci of sclerosis was diagnosed at core biopsy in two lesions in which histology showed papillary lesion associated with carcinoma in situ. Therefore, core biopsy had a sensitivity of 82% (95% CI, 59-100%), specificity of 100% (100-100%), positive predictive value of 100% (100-100%), and negative predictive value of 83% (62-100%) for the detection of malignancy in papillary lesions (Table 1).
|
Correlation of Imaging Features with Mastectomy or Excision Results
Only eight patients underwent galactography. Intraluminal filling defects
were identified in seven ductograms. One ductogram failed because of high
pressure encountered during the injection of contrast material, as a result of
which the duct could not be well delineated.
Thirty-five lesions (62.5%) were not detected on mammography. Of the 21 lesions detected on mammography, 10 were ductal carcinoma in situ, three were papillary carcinoma, six were papilloma, and two were sclerosed papilloma. The mammographic features of these lesions are listed in Table 2. Using the characteristics of the mass shown on mammography to differentiate malignant from benign papillary lesions would therefore give a sensitivity of 67%, positive predictive value of 50%, and both specificity and negative predictive value of less than 10%. Characteristics of microcalcifications (Fig. 1) used to predict malignancy gave a sensitivity of 75%, specificity of 50%, positive predictive value of 75%, and negative predictive value of 25%. A total of nine malignant lesions showed at least one or more suspicious mammographic features; four of the malignant lesions did not show any suspicious malignant features. Of the eight benign lesions, six showed one or more suspicious malignant lesions. The overall sensitivity was therefore 69% (95% CI, 44-94%); specificity, 25% (-5% to 55%); positive predictive value, 60% (35-85%); and negative predictive value, 33% (-4% to 71%) (Table 2).
|
|
Sonography was more sensitive than mammography in the detection of abnormality in papillary lesions in the breast. Abnormality in sonography was detected in 46 lesions (82.1%). Ten papillomata were not detected sonographically, of which five were coincident histologic findings found in the mastectomy specimens and five were diagnosed at segmental duct resection. The sonographic features of the 25 papillomata, one atypical papilloma, four sclerosed papillomata, and 16 malignant lesions are presented in Table 3 and shown in Figures 2 and 3. Of all 16 malignant lesions, nine showed one or more sonographically suspicious features and the remaining seven were benign-appearing on sonography. Of the 30 benign lesions, three showed one or more suspicious sonographic features and the remaining 27 appeared benign on sonography. Sonography had a sensitivity of 56% (95% CI, 32-81%), specificity of 90% (79-100%), positive predictive value of 75% (51-100%), and negative predictive value of 79% (66-93%).
|
|
|
Twenty-one lesions were detected on both mammography and sonography. When the presence of any suspicious mammographic or sonographic feature was considered to indicate malignancy, interpretation of both mammographic and sonographic features gave a sensitivity of 61% (95% CI, 39-84%), specificity of 33% (-20% to 87%), positive predictive value of 85% (65-100%), and negative predictive value of 13% (-10% to 35%) for the detection of malignancy in papillary lesions. The combined interpretation of both mammography and sonography therefore increased the positive predictive value but the diagnosis became less specific.
Correlation of Both Imaging Features and Core Biopsy Results with Histologic Findings
Of all the 21 lesions having core biopsy, 13 had concordant imaging and
biopsy findings. Of these 13 lesions, seven were benign and six were
malignant. One of these lesions with concordant imaging and core biopsy, which
suggested a benign abnormality, eventually showed a malignant lesion on
excision. The histologic findings agreed with the biopsy and imaging findings
in the remaining 12 lesions.
Eight lesions had discordant imaging and biopsy findings. Four of these eight lesions were malignant histologically. The imaging features were suspicious of malignancy in only one of these lesions. The remaining four histologically proven benign lesions were all benign at core biopsy but had suspicious imaging features.
Correlation of Both Imaging Features and Fine-Needle Aspiration Biopsy Results with Histologic Findings
Of all the 37 lesions having fine-needle aspiration biopsy, 22 had
concordant imaging and fine-needle aspiration findings. Of these 22 lesions,
20 were benign and two were malignant histologically. Three lesions with
concordant imaging and fine-needle aspiration findings suggestive of
malignancy were proven to be benign histologically. None of the malignant
lesions showed concordant benign imaging and fine-needle aspiration findings
in this small series.
Fifteen lesions had discordant imaging and fine-needle aspiration findings. Eight of these lesions were benign and seven were malignant.
Clinical Follow-Up
One patient developed a papilloma in the contralateral breast 1 year after
the excision of the presenting papilloma. Another patient developed carcinoma
in situ in the ipsilateral breast 1 year after the excision of the presenting
papilloma. Three of 13 patients developed papillary carcinoma in the
ipsilateral breast 1-2 years after the initial diagnosis of carcinoma in situ
by surgical excision.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Solitary breast papillomata are potentially malignant and are associated with a high risk of breast cancer [5]. In our series, up to 12 patients (30%) who had carcinoma in situ or carcinoma had papilloma before or after the diagnosis of a presenting carcinoma. Such a high association between papilloma and breast cancer has also been reported previously [5]. Gutman et al. [5] reviewed 95 women with papilloma or papillomatosis and found that in up to 10% of patients, solitary papillomata were associated with breast carcinoma. An additional 9% of patients presented with invasive or noninvasive carcinoma in the papilloma. Patients with papilloma also have a high chance of developing a second papilloma. MacGrogan and Tavassoli [6] found that the presence of epithelial hyperplasia, ductal hyperplasia, or lobular neoplasia in the surrounding breast as well as infarction of the papilloma were significant predictive factors of recurrence. Because only one patient in our series had recurrent papilloma, we could not identify any factors for the prediction of recurrence.
Radiologically, the lesion can be small and can be mammographically and sonographically occult. The overall sensitivity for detection of papillary lesions on mammography is low (37.5% in our series). In the series by Woods et al. [7], abnormal mammographic findings were seen in eight of 19 patients, and the incidence of abnormal mammographic findings was similar to that in our series. Even when the lesion is detectable on mammography, such detection is neither sensitive nor specific enough for accurate differentiation between malignancy and benignancy. The presence of calcification in papilloma was rare in our series: only two lesions showed punctate microcalcification [8, 9]. A higher proportion of malignant lesions showed associated microcalcification (62.5%). When the lesion is detectable on mammography, analysis of the characteristics of microcalcifications is slightly more sensitive than analysis of the characteristics of the mass in differentiating between benign and malignant lesions. When patients present with nipple discharge, galactography is a sensitive [7, 9] but nonspecific method [10] for the detection of intraductal lesions. A high incidence of abnormal findings on galactography was also shown in our small series. Unfortunately, many of our patients did not present with nipple discharge so the role of galactography was limited in our study.
The presence of dilated ducts is a common sonographic finding that is often associated with a visible intraluminal echo [8]. We also found dilated ducts associated with 10 of our sonographically detectable lesions (21.7%). A dilated duct is a sonographic sign useful for differentiating intraductal lesions from other benign lesions such as fibroadenoma. Sonographically, papillomata usually present as a circumscribed hypoechoic oval nodule. Vascularity identified in the fibrovascular core of the papilloma might suggest the diagnosis. Color Doppler images were unfortunately not available for all lesions and most of the papillomata in our series were small; vascularity is therefore not shown in our series. Differentiating sclerosed papilloma from papillary carcinoma is also difficult. In our series, one of four sclerosed papillomata had a microlobulated border and another one was complex in echogenicity.
The margin of the lesion and the echo pattern appeared to be useful sonographic signs for differentiating benign and malignant lesions. Although these signs are fairly specific, the low sensitivity suggests that a number of malignant lesions might be missed if one relies only on imaging features to identify these lesions. Combining the sonographic and mammographic findings does not make the diagnosis more sensitive or specific. As a result, one cannot reliably diagnose a papillary lesion as benign or malignant simply on the basis of imaging features, nor could one use this assessment to decide whether to perform biopsy.
Michael and Buschmann [11] reviewed 22 fine-needle aspirations from histologically proven papillary neoplasms (10 papillary carcinomas and 12 intraductal papillomata) and concluded that papillary breast neoplasms can be accurately classified by cytology. Gomez-Aracil et al. [12] also suggested that papillary carcinoma of the breast can be diagnosed on cytology and differentiated from papilloma. However, the diagnosis of papillary carcinoma can be difficult by fine-needle aspiration. Fine-needle aspiration might miss the small foci of carcinoma in situ or that are invasive.
A similar limitation might be encountered at core biopsy [2, 13, 14]. Some authors have suggested that benign papillomata diagnosed at core biopsy can be followed up clinically and radiologically so that surgical excision is not necessary [14]. That a papillary carcinoma in situ diagnosed at core biopsy might be upgraded to invasive carcinoma when the entire lesion is examined is a significant limitation of core biopsy in the management of papillary lesions [15-17]. In our series, again both fine-needle aspiration and core biopsy showed their limitations for the diagnosis of papillary lesions.
Considering these limitations, one might question whether all papillary lesions, including benign papillomata diagnosed at core biopsy or fine-needle aspiration, should be surgically excised. The current practice is not to excise benign papillomata diagnosed at needle biopsy if there is no atypia and no discordance between imaging and histologic findings. In our small series, one lesion showing both benign radiologic features and benign core biopsy results was found to be ductal carcinoma in situ at histology. Our results also showed that a significant proportion of malignant papillary lesions did not show any malignant features mammographically or sonographically. Our small sample size suggests that even concordance between imaging and biopsy findings might fail to detect malignant papillary lesions.
The use of larger-gauge needles would allow larger cores of tissue and more accurate histologic assessment. Recent studies recommended the use of stereotactic vacuum-assisted biopsy for the diagnosis of papillary lesions [18]. We were not able to analyze the role of stereotactic vacuum-assisted biopsy in our study because that procedure has been used in our center for only the past 2 years.
A significant limitation of our study is that most of our core biopsies consisted of three cores obtained with a 16-gauge needle, which is considered undersampling by present standards. The small gauge of the needle may have a significant impact on the results. However, even when stereotactic vacuum-assisted biopsy is used, it is not 100% accurate in the histologic diagnosis of papillary lesions [19]. Mercado et al. [19] therefore suggested that the follow-up period for patients with papillary breast lesions should be extended to more than 2 years [19]. Excision biopsy to rule out malignancy was also recommended when atypical hyperplasia of the lesion was shown at stereotactic vacuum-assisted biopsy [19].
In conclusion, we analyzed the imaging features of papillary lesions according to the BI-RADS and BI-RADS US lexicon classifications, and we found imaging features such as positive findings on galactography and associated ductal dilatation to be useful signs in establishing the diagnosis of papillary lesion. However, neither mammographic nor sonographic features are sensitive and specific enough to allow accurate differentiation between benign and malignant lesions. The limitations of fine-needle aspiration and core biopsy should be kept in mind. Papillary lesions, with their high association with breast cancer, should be closely monitored in all patients. Excisional biopsy or stereotactic vacuum-assisted biopsy should be used to evaluate these lesions. Multiple lesions found in both breasts might create difficulty in management. If excisional biopsy or stereotactic vacuum-assisted biopsy of all papillary lesions is not feasible, serial follow-up and intense scrutiny are prudent.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
This article has been cited by other articles:
|
|
 |
|
  S. A. Schwab, M. Uder, R. Schulz-Wendtland, W. A. Bautz, R. Janka, and E. Wenkel Direct MR Galactography: Feasibility Study Radiology, October 1, 2008; 249(1): 54 - 61. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. H. Shim, E. J. Son, E.-K. Kim, J. Y. Kwak, J. Jeong, and S. W. Hong Benign Intracystic Papilloma of the Male Breast J. Ultrasound Med., September 1, 2008; 27(9): 1397 - 1400. [Full Text] [PDF]
|
|
|
|
 |
|
 M. Muttarak, P. Lerttumnongtum, B. Chaiwun, and W. C. G. Peh Spectrum of Papillary Lesions of the Breast: Clinical, Imaging, and Pathologic Correlation Am. J. Roentgenol., September 1, 2008; 191(3): 700 - 707. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
H. J. Shin, H. H. Kim, S. M. Kim, H. R. Yang, J.-H. Sohn, G. Y. Kwon, and G. Gong Papillary Lesions of the Breast Diagnosed at Percutaneous Sonographically Guided Biopsy: Comparison of Sonographic Features and Biopsy Methods Am. J. Roentgenol., March 1, 2008; 190(3): 630 - 636. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
T. H. Kim, D. K. Kang, S. Y. Kim, E. J. Lee, Y. S. Jung, and H. Yim Sonographic Differentiation of Benign and Malignant Papillary Lesions of the Breast J. Ultrasound Med., January 1, 2008; 27(1): 75 - 82. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 V. Sohn, J. Keylock, Z. Arthurs, A. Wilson, G. Herbert, J. Perry, M. Eckert, D. Smith, S. Groo, and T. Brown Breast Papillomas in the Era of Percutaneous Needle Biopsy Ann. Surg. Oncol., October 1, 2007; 14(10): 2979 - 2984. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
T. Rissanen, H. Reinikainen, and M. Apaja-Sarkkinen Breast Sonography in Localizing the Cause of Nipple Discharge: Comparison With Galactography in 52 Patients J. Ultrasound Med., August 1, 2007; 26(8): 1031 - 1039. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
