Wo m e n ’s I m a g i n g • O r i g i n a l R e s e a r c h Harvey et al. Follow-Up Imaging of Palpable Breast Lesions Women’s Imaging Original Research

W O M E N ’S IMAGING

Jennifer A. Harvey 1 Brandi T. Nicholson1 Alexander P. LoRusso1,2 Michael A. Cohen1 Viktor E. Bovbjerg 3,4 Harvey JA, Nicholson BT, LoRusso AP, Cohen MA, Bovbjerg VE

Keywords: breast neoplasms, mammography, ultrasound DOI:10.2214/AJR.09.2811 Received March 26, 2009; accepted after revision May 21, 2009. 1 Department of Radiology, University of Virginia, Box 800170, Charlottesville, VA 22908. Address correspondence to J. A. Harvey. 2

Present address: X-Ray Associates of New Mexico, Santa Fe, NM.

3 Department of Public Health Sciences, University of Virginia, Charlottesville, VA. 4

Present address: College of Health and Human Sciences, Oregon State University, Corvallis, OR.

AJR 2009; 193:1723–1730 0361–803X/09/1936–1723 © American Roentgen Ray Society

AJR:193, December 2009

Short-Term Follow-Up of Palpable Breast Lesions With Benign Imaging Features: Evaluation of 375 Lesions in 320 Women OBJECTIVE. The purpose of this study was to evaluate the feasibility of short-term follow-up of palpable masses that have benign imaging features. MATERIALS AND METHODS. The cases of all women with round, oval, or lobular palpable masses with circumscribed margins and homogeneous ultrasound echotexture for which short-term follow-up was recommended from July 1997 through December 2003 were retrospectively identified. Evaluation was by ultrasound and/or mammography and focused clinical examination. Outcome was assessed with imaging or clinical follow-up lasting at least 12 months. The cancer incidence for palpable lesions was compared with that for nonpalpable lesions recommended for short-term follow-up. RESULTS. In 379 women, 443 palpable masses with benign features for which shortterm follow-up was recommended were identified. Outcome data were available on 375 masses in 320 women. Lesions were evaluated with mammography and ultrasound (n = 186) or ultrasound alone (n = 189). Masses were typically identified only with ultrasound (n = 258, 68.8%); were oval (n = 275, 73.3%), of equal density to normal breast tissue on mammograms (n = 95 on 117 mammograms, 81.2%), and hypoechoic (n = 336 in 372 ultrasound examinations, 90.3%); and were prospectively believed to be fibroadenoma (n = 304, 81.1%). Eightyfive lesions (22.7%) were biopsied soon after evaluation, and one 1.5-mm ductal carcinoma in situ was diagnosed. At follow-up (mean, 2.7 years), 26 lesions (6.9%) had grown. Twentyfour of the 26 lesions were biopsied, and no cancer was diagnosed. The overall cancer prevalence was similar for palpable (0.3%) and nonpalpable (1.6%) masses. The cost of short-term follow-up was less than that of biopsy. CONCLUSION. Short-term follow-up is a reasonable alternative to biopsy of palpable breast lesions with benign imaging features, particularly for young women with probable fibroadenoma.

T

he indications for short-term follow-up of nonpalpable breast lesions are well established. BIRADS calls for initial short-term follow-up (BI-RADS assessment category 3, probably benign finding) of solid well-defined masses, focal asymmetries, and grouped punctate calcifications on a baseline mammogram or when previous mammograms cannot be obtained [1]. In the early 1990s, Sickles [2] and Varas et al. [3] found short-term follow-up of these lesions a reasonable alternative to biopsy because the incidence of breast cancer was less than 2% and cancer subsequently diagnosed was at an early stage. Cancer incidence later was reported to be independent of lesion size and patient age [4]. Although Sickles [2] did not explicitly state that women with palpable findings were

excluded from the study, later clarification confirmed that they had been [5]. Of note, the third edition of BI-RADS [6], published in 1998, did not address whether in the assignment of assessment category 3, probably benign finding, lesion management would differ depending on the palpable or nonpalpable nature of the lesion. Regarding the use of BI-RADS category 3, the third edition stated, “At the present time, most approaches are intuitive. These will likely undergo future modification as more data accrue as to the validity of an approach, the interval required, and the type of findings that should be followed.” It was not until publication of the fourth edition of BI-RADS [1] in 2003 that a distinction was made for management of palpable lesions. The fourth edition of BIRADS states that “all published studies ex-

1723

Harvey et al. clude palpable lesions, so the use of a probably benign assessment for a palpable lesion is not supported by scientific data.” The criteria for short-term follow-up of masses with benign features at ultrasound examination are not as well validated as they are for such masses found at mammography. Using sonographic criteria, Stavros et al. [7] described benign and malignant features of 750 solid masses. Lesions classified prospectively as benign had a cancer incidence of 0.5%. Of interest, 278 women (37%) in the total study sample of 750 in that study had palpable findings. The study outcomes were similar for palpable and nonpalpable lesions. More recently, Graf et al. [8] found a 0.2% incidence of cancer (95% CI, 0.0–1.23%) among 445 solid nonpalpable masses with benign features detected with ultrasound that had been obscured or partly obscured at mammography. The ultrasound criteria for inclusion in the study were oval or lobular shape, circumscribed margins, parallel orientation, isoechoic or mildly hypoechoic echotexture, and absence of shadowing. Subsequent studies have well documented the outcome of short-term follow-up of nonpalpable lesions with benign features on baseline images [9, 10], but little information exists on palpable lesions. In 2004, Graf et al. [11] described a series of 157 palpable noncalcified solid masses with benign imaging features that were observed with short-term follow-up rather than biopsied. No cancer was diagnosed among the palpable lesions in that study. All patients in the study underwent mammography with or without ultrasound and were observed for at least 2 years (mean, 4.1 years). Owing to the findings of early studies of short-term follow-up as an alternative to biopsy, in the mid-1990s all faculty members at our institution began recommending short-term follow-up of masses that had benign features at imaging but did not have ultrasound features of simple cyst. Both palpable and nonpalpable lesions were included. In this study, we retrospectively evaluated our experience with short-term follow-up of 375 palpable lesions with benign imaging features in 320 consecutively registered women. Materials and Methods This retrospective study was approved by our institutional review board for health sciences research. A waiver of consent was granted. This study was compliant with HIPAA.

1724

Patient Selection Using our computerized reporting system, we identified the cases of all women with palpable lesions described as round, oval, or lobular masses with circumscribed margins for which short-term follow-up had been recommended between July 1997 and December 2003. Lesions with other imaging features, such as irregular shape or ill-defined margins, were excluded because short-term followup of these lesions, even if nonpalpable, would be controversial. Women were excluded if the recommendation was based on interpretation of images from another institution or if follow-up was recommended because of benign needle biopsy findings.

nant mass, regardless of imaging features. Four of the six radiologists providing services during the study period had formal training in clinical breast examination through a breast and cervical cancer screening program or during their breast imaging fellowship. The other two radiologists had no training other than what they received during medical school and internship. There was no fee or report for the clinical examination because it was focused only on the area of concern by the patient or referring physician. Focused clinical breast examination was performed at each shortinterval follow-up imaging visit.

Patient Care Imaging and Image Interpretation Women underwent evaluation with mammography, ultrasound, or both. Our clinical practice is generally to use ultrasound as the initial technique for evaluation of palpable lesions in women younger than 30–35 years, pregnant women, and women with normal findings on a screening mammogram within the previous 4–6 months. The choice to use ultrasound or mammography for evaluation of women 30–35 years old is at the radiologist’s discretion. Women older than 30–35 years without normal findings on a recent screening mammogram typically undergo both mammography and ultrasound. All images were interpreted by one of six radiologists who met the qualifying criteria of the Mammography Quality Standards Act. The radiologists’ length of experience in breast imaging varied from 9 months to 10 years. Short-term follow-up was recommended for round, oval, or lobular masses with circumscribed margins on mammograms or ultrasound images. Lesions identified on ultrasound images had homogeneous echotexture and no malignant features. Lesions that were characteristically benign, such as simple cysts, oil cysts, and lymph nodes, were assessed BI-RADS category 2, benign finding, and were not included in this study. A statement of the condition likely represented by the finding, such as probable fibroadenoma or complicated cyst, was usually included in the report. Images were compared with older images when available. Short-term follow-up was not recommended for lesions that were new compared with findings of previous breast imaging studies.

Clinical Examination All women underwent a correlative focused clinical breast examination of the palpable finding by a radiologist, as is our standard practice for all women with palpable findings. Short-term follow-up was not performed when the clinical examination finding was suspicious, such as when the palpable finding represented a hard domi-

Because the palpable lesions were both imaging and clinical findings, assessment of stability was established with either imaging or clinical examination. Primary review in this retrospective study was based on outcome on subsequent breast imaging reports. If no biopsy or subsequent imaging was performed or if imaging follow-up lasted less than 12 months, medical records were reviewed. Clinical notes were reviewed to assess whether a breast examination had been performed that had normal findings or showed no change in the palpable finding at least 12 months after the initial assessment or whether breast cancer had been diagnosed. Lesion outcome was assessed as one of the following: tissue diagnosis; decrease in lesion size or level of suspicion (e.g., the lesion was clearly an oil cyst on the 6-month mammogram) at imaging follow-up and lesion reported as BI-RADS category 2, benign finding; lesion resolution at imaging follow-up and finding reported as BI-RADS category 1, negative; or lesion stability for 12 months or longer on images or lesion stability for 12 months or longer at clinical examination according to review of medical records. Lesions subjected to tissue diagnosis rather than the recommended short-term follow-up because of patient preference, referring health care provider request, or other reason, such as presence of ipsilateral breast cancer, were included in the analysis because short-term followup had been recommended.

Statistical Analysis Cancer outcome was determined for palpable lesions for which short-term follow-up was recommended. Comparison was made with the cancer outcome among nonpalpable breast masses with a BI-RADS category 3 assessment from January 1998 through December 2000. This time period was used because the data had been previously collected for another institutional review board–approved study and were readily available for comparison. The cost of follow-up compared with immediate excisional or core needle biopsy was evaluat-

AJR:193, December 2009

Follow-Up Imaging of Palpable Breast Lesions ed with current Medicare reimbursement values. The current rates are $86.06 for breast ultrasound, $81.68 for unilateral diagnostic mammography, $103.39 for bilateral diagnostic film-screen mammography, $2,174 for excisional biopsy, and $617.45 for ultrasound-guided core needle biopsy. The cost of biopsy of lesions that increased in size or level of suspicion was taken into account. Exact binomial confidence intervals (CIs) were calculated to compare the cancer incidence among women with palpable BI-RADS category 3 breast masses with the incidence among women with nonpalpable BI-RADS category 3 breast masses.

Results The cases of 462 women with 537 palpable lesions for which short-term follow-up was recommended were identified between July 1997 and December 2003. Eighty-three women with 94 lesions were excluded because the lesions described were other than round, oval, or lobular or had other than circumscribed margins. These lesions included probable abscess, sebaceous cyst, and scar. Twenty women with 20 lesions were excluded because neither imaging nor clinical follow-up findings were available. Thirty-nine women with 48 lesions were excluded because imaging or clinical follow-up showed the lesion had been stable for fewer than 12 months. The final study group therefore included 320 women with 375 palpable masses with benign features for which short-term follow-up was recommended. The mean patient age was 34.4 years (range, 12.2–87.8 years; median, 32.8 years). Lesions were evaluated with mammography and ultrasound (n = 186, 49.6%) or ultrasound alone (n = 189, 50.4%). Of the 189 women who underwent evaluation with ultrasound only, 151 were 30 years old or younger, and 20 were 31–35 years old. All 15 women evaluated with ultrasound only who were 40 years old and older had had normal findings at screening mammography within the previous 4 months. Two women, both 36 years old, had not undergone mammography at evaluation or within the previous year. Mammography was deferred for a 36-yearold pregnant woman with a probable lactational adenoma. Most of the lesions were identified with ultrasound alone (n = 258, 68.8%) or with mammography and ultrasound (n = 114, 30.4%). Only three lesions (0.8%) were identified with mammography alone. The mean mass size was 14 mm (range, 2–98 mm; median, 13 mm). Mass size was not reported

AJR:193, December 2009

TABLE 1: Characteristics and Assessment of Masses with Benign Features (n = 375) Mass Characteristic

No.

Shape Oval



275 (73.3)

Round



49 (13.1)

Lobular



51 (13.6)

Equal



95 (81.2)

High



2 (1.7)

Not specified



20 (17.1)

Hypoechoic



336 (90.3)

Isoechoic



34 (9.1)

Hyperechoic



2 (0.5)

Density on mammogram (n = 117)

Echotexture on ultrasound image (n = 372)

Prospective lesion assessment Fibroadenoma



304 (81.1)

Complicated cyst vs solid mass



66 (17.6)

Not given



5 (1.3)

Note—Values in parentheses are percentages.

for 19 lesions. Masses were most commonly oval (n = 275, 73.3%), equal in density to surrounding normal breast tissue on the 117 mammograms (n = 95, 81.2%) and hypoechoic at the 372 ultrasound examinations (n = 336, 90.3%) (Table 1). Lesions were most frequently thought to be probable fibroadenoma or complicated cyst as opposed to a solid mass (Table 1). Specifically, 304 masses (81.1%) in 262 women were reported as probable fibroadenoma (Fig. 1). Women with probable fibroadenoma were younger (mean age, 32.0 years; range, 12.2–87.8 years; median, 30.9

years) than women with a complicated cyst as opposed to a solid mass (mean, 47.0 years; range, 15.0–75.3 years; median 47.6 years). As expected, the peak prevalence of probable fibroadenoma occurred among women 20–40 years old (Table 2). Masses assessed as probable fibroadenoma were more frequently multiple in women younger than 30 years (Table 2). Most of the palpable lesions (n = 264, 70.4%) evaluated with follow-up had a benign outcome (Table 3). These lesions either resolved (n = 86, 22.9%), decreased in size or level of suspicion with a change in BI-RADS

Fig. 1—28-year-old woman with palpable mass in right breast. Ultrasound image shows corresponding 32-mm oval circumscribed hypoechoic solid mass (arrows) with orientation parallel to chest wall. Typical case for followup. Mass contains internal septation (arrowhead). Appearance and clinical presentation strongly suggest fibroadenoma. Mass was stable at follow-up for more than 5 years.

1725

Harvey et al. TABLE 2: Age Distribution of Masses Assessed as Probable Fibroadenoma No. of Women With Multiple Lesionsb

No. of Women (n = 262)a

No. of Lesions (n = 304)a

12–19



29 (11.1)



37 (12.2)



8 (27.6)

20–29



88 (33.6)



109 (35.9)



11 (12.5)

30–39



81 (30.9)



88 (28.9)



6 (7.4)

40–49



43 (16.4)



47 (15.5)



4 (9.3)

50 and greater



21 (8.0)



23 (7.6)



1 (4.8)

Age (y)

aValues in parentheses are percentages.

bValues in parentheses are percentage of women in age group with probable fibroadenoma.

TABLE 3: Outcome of 375 Palpable Lesions in 320 Women Lesion Outcome

No.

No. of Cases of Breast Cancer

Sampled because of patient or health care provider preference or other reason

85 (22.7)

1

Resolved (BI-RADS 1)

86 (22.9)

0

Decreased in size or level of suspicion (BI-RADS 2)

65 (17.3)

0

Stable (BI-RADS 2) (mean follow-up, 30 mo; range, 12–94 mo)

113 (30.1)

0

Increase in size or level of suspicion at follow-up

26 (6.9)

0 (24/26 sampled)

Note—Values in parentheses are percentages.

assessment category to 2 (benign finding) (n = 65, 17.3%), or were stable at follow-up (n = 113, 30.1%). An example of a lesion that decreased in level of suspicion was a mass that was oval, circumscribed, and equal-density at initial imaging that was clearly a benign lymph node or oil cyst at follow-up imaging.

Of the 113 lesions that were stable at followup for at least 12 months, 84 were stable at imaging and 29 were clinically stable according to medical record review. The overall mean follow-up period was 2.7 years (range 1.0–8.3 years; median, 2.1 years). For the lesions evaluated with clinical follow-up,

the mean follow-up period was 3.6 years (range 1.0–8.3 years; median, 3.3 years). Eighty-five lesions (22.7%) were evaluated with tissue diagnosis instead of or during follow-up because of the preference of the patient, health care provider request, or other reason, such as presence of ipsilateral breast carcinoma (Table 3). Twenty-six lesions (6.9%) had increased in size or level of suspicion at imaging follow-up. Biopsy was recommended for all 26 lesions, but one patient was lost to follow-up and one patient declined biopsy. Of the 24 lesions that had known pathologic outcomes available, none was cancerous. One case of cancer was diagnosed among the 375 palpable lesions (0.3%) for which short-term follow-up had been recommended (Table 3). The patient had a 2.5-mm round circumscribed hypoechoic mass that was considered to be either a cyst or solid mass (Fig. 2). Surgical biopsy was performed at the patient’s request. A 1.5-mm focus consisting of two involved ducts of cribriform ductal carcinoma in situ (DCIS) was found. Review of the histologic findings predominantly showed fibrocystic changes, which likely accounted for the hypoechoic mass seen at ultrasound imaging, rather than the sparse focus of DCIS. This case was included in this study as a false-negative worstcase scenario, although it is likely that the DCIS found in this case was an incidental

A

B

Fig. 2—59-year-old woman with palpable lump in 12 o’clock position of left breast. A, Craniocaudal spot compression mammogram of left breast shows no abnormality. Triangle represents adhesive marker on skin to designate location of palpable finding. B, Sonogram of left breast shows 2.5-mm round circumscribed hypoechoic mass (arrow) that is not clearly simple cyst. This was smallest lesion in study but was considered to correspond to palpable clinical finding. Patient chose surgical excision, at which 1.5-mm focus of cribriform ductal carcinoma in situ in background of fibrocystic changes was identified.

1726

AJR:193, December 2009

Follow-Up Imaging of Palpable Breast Lesions finding and not related to the palpable or imaging finding. There was no statistically significant difference between palpable and nonpalpable lesions recommended for short-term follow-up at our institution (Table 4). The 0.3% incidence of malignancy of palpable masses with benign features in this study was lower than the cancer rate of 1.6% for nonpalpable masses designated probably benign at our institution, although the difference was not statistically significant (p = 0.063, Fisher’s exact test) (Table 4). One hundred nine lesions were subjected to tissue sampling (85 because of patient or health care provider preference, 24 because of an interval increase in size or suspicion). Fifty-eight of the sampled lesions (53.2%) were fibroadenoma (n = 55, 50.5%), lactational adenoma (n = 2), or tubular adenoma (n = 1), all of which have similar imaging findings. An additional fibroadenoma contained atypical ductal hyperplasia at core needle biopsy and was surgically excised; no malignancy was identified. Twenty-four lesions were cysts (n = 13, 11.9%) or fibrocystic change (n = 11, 10.1%). One patient each had lobular carcinoma in situ and atypical ductal hyperplasia; both lesions were excised and no malignancy was found. A variety of other pathologic findings were made at biopsy, including fibrosis, papilloma, hyperplasia, nodular adenosis, and juvenile hypertrophy. Twenty of the 24 lesions sampled because of increase in size or suspicion at follow-up were prospectively considered probable fibroadenoma. Among the 20 patients with these 20 probable fibroadenomas, surgical pathologic results were fibroadenoma in 14 patients, lactational adenoma in two patients, and tubular adenoma in one patient. Thus 85% (17 of 20) of the masses prospectively classified as probable fibroadenoma that had increased in size or suspicion at follow-up were correctly assessed, because lactational adenoma and tubular adenoma have the same appearance as fibroadenoma [12]. The pathologic findings at examination of the biopsy specimens from the other three patients were adenolipoma, lobular atrophy, and nodular adenosis. The cost of short-term follow-up was lower than that of immediate biopsy with either ultrasound-guided core or excisional technique (Table 5). Medicare reimbursement for breast ultrasound is similar to that of a unilateral film-screen mammogram ($82), but lower than a unilateral digital mammogram ($117). The cost is lower for women 40 years and older with a lesion seen at mammography because mammography would be performed at annual

AJR:193, December 2009

TABLE 4: Comparison of Palpable and Nonpalpable Breast Masses With Recommendation of Short-Term Follow-Up Nature of Lesion

Time Period

Cancer Incidence

95% CI (%)

Palpable

7/97–12/03

1/375 (0.3%)

0.0001–0.015

Nonpalpable

1/98–12/00

10/641 (1.6%)

0.8–2.9

TABLE 5: Cost Comparison Between Follow-Up Imaging and Immediate Biopsy Procedure

Medicare Reimbursement ($)

Woman with lesion seen on ultrasound imaging­— follow-up Ultrasound at 6 mo

86

Ultrasound at 12 mo

86

Ultrasound at 24 mo

86

Excisional biopsy for 7% of women because of interval growth

152

Total

410

Woman age 40 y or older with lesion seen on mammographya—follow-up Unilateral mammogram at 6 mo

82

Excisional biopsy for 7% of women because of interval growth

152

Total

234

Immediate ultrasound-guided core needle biopsy Ultrasound-guided core needle biopsy Ultrasound 6 mo after benign biopsy finding Ultrasound 12 mo after benign biopsy finding Total Excisional biopsy

617 86 86 789 2,174

aWomen age 40 and older are assumed to undergo bilateral mammography as part of ongoing clinical

care.

intervals anyway. The cost for these women would be only the cost of the 6-month followup mammogram. The costs are slightly higher with the use of digital mammography (about $35 per study). The cost of biopsy because of interval growth of some lesions was factored into the evaluation by multiplying the cost of excisional biopsy ($2,174) by 7%. Medicare reimbursement varies by region. Our results are based on reimbursement in our region but would not likely be significantly different in other regions. Discussion The use of short-term follow-up for management of a noncalcified well-circumscribed solid mass found on a baseline screening mammogram is well accepted when the mass is not palpable. These lesions have benign characteristics and are considered BI-RADS category 3, probably benign finding. The likelihood that these lesions are carcinoma is small, less than 2% risk, and cancer is typically at an early stage when subsequently diagnosed.

Our study of palpable breast lesions with benign features showed an acceptably low prevalence of breast cancer (0.3%), so low that shortterm follow-up is a reasonable alternative to biopsy. The prevalence of breast cancer among the palpable masses with benign features in our study was lower than that among nonpalpable lesions both at our institution and documented in the literature [2, 3, 9]. Overlap of the 95% CIs of palpable and nonpalpable masses indicates that the lower incidence of cancer among palpable lesions is not statistically different. In studies of short-term follow-up of probably benign lesions, Sickles [2] and Varas et al. [3] found the risk of cancer was highest among nonpalpable noncalcified well-defined solid masses. Sickles found 12 cases of cancer among 589 masses (2.0%), and Varas et al. found four cases of cancer among 289 masses (1.4%). Our experience was similar: a 1.6% incidence of cancer among nonpalpable masses classified BI-RADS category 3. A mass may be detected as a palpable finding owing to interval growth, recent initiation

1727

Harvey et al. of clinical or breast self-examination, or overall softening of the breast texture over time. One might argue that the presence of a palpable finding at examination indicates that the lesion is clinically growing, necessitating biopsy. However, patients often are not familiar with breast self-examination and may detect masses that have been present for many years. We have had patients who report a “new” palpable lump that represents a degenerating fibroadenoma or oil cyst that has been stable on images for many years. These are long-standing lesions, and it is not known why women detect them at any particular point in time. The breast may be undergoing involution, or the patient may have lost weight, which results in an improved ability to palpate a lesion, but usually there is no apparent explanation. In addition, some women who undergo evaluation of a palpable breast mass know that it has been present and stable for many years, but they have seen a new physician who is concerned about the finding. Of interest is that only 6.9% of the lesions in our study had a follow-up finding showing an increase in size, suggesting that most of these masses probably had not been detected at examination owing to a change in size. The use of short-term follow-up of nonpalpable lesions with benign features is well accepted in clinical practice. Follow-up of palpable masses may be more reliable than that of nonpalpable masses because lesion change can be readily assessed with physical examination as well as with imaging. It is possible for a nonpalpable mass to double in size and the change not be detected until subsequent imaging is performed. When a lesion is palpable, however, patients and referring physicians can readily assess for change in size at clinical examination. Our patients are instructed to return promptly if they are concerned that a palpable finding has increased in size. Some patients may desire biopsy of a palpable finding because of anxiety. In our study 22.7% of patients chose biopsy rather than follow-up. This percentage is higher than the 0.5–0.9% biopsy rate in studies of nonpalpable lesions for which short-term follow-up is recommended [2, 10]. The presence of a palpable lump can cause more anxiety than a nonpalpable finding does. In our practice, if a patient appears to be anxious during discussion of imaging results, biopsy is offered. Most patients, however, choose short-term follow-up. Our study also showed that shortterm follow-up is less expensive than immedi-

1728

ate biopsy. We did not, however, consider the cost of a delay in diagnosis and treatment because the only malignant tumor in our study was a very small focus of DCIS that was diagnosed soon after the initial imaging. Palpable breast masses most commonly are fibroadenomas in women in the second and third decades of life, cysts or fibrocystic changes in women in the fourth decade, and cancer in women older than 50 years [13]. The one case of cancer in our series occurred in a 59-year-old woman. Most of the masses followed up in our study were prospectively classified as probable fibroadenoma (81.1%), oval (73.3%) with circumscribed margins, equal mammographic density (81.1%), and hypoechoic (90.3%). Fibroadenoma is more common among younger women; the mean age of the women in our study with probable fibroadenoma was 32.0 years. Skaane and Engedal [14] found a negative predictive value of 100% for palpable and 96% for nonpalpable lesions with the following six grouped sonographic features: round or oval presence of a pseudocapsule; homogeneous echotexture; high, equal, or slightly low echogenicity; no shadowing; and normal surrounding tissue. This spectrum of imaging findings suggests fibroadenoma. Fibroadenoma can fluctuate in size, prompting biopsy for some patients. Gordon et al. [15] confirmed that growth of fibroadenoma, particularly in premenopausal women, is common, occurring in 194 of 1,070 fibroadenomas (18%) diagnosed with fine-needle aspiration biopsy. They further found that fibroadenoma diagnosed with fine-needle aspiration biopsy can grow by as much as 16% of its volume per month if the patient is younger than 50 years and as much as 13% if the patient is 50 years or older. The interval growth of lesions in our study was lower (6.9%), but our study included all masses with benign features, not known fibroadenoma alone. A concern about recommending shortterm follow-up of palpable solid masses with benign features may be the difficulty in differentiating fibroadenoma from phyllodes tumor because these two lesions have similar features on mammograms and ultrasound images. Phyllodes tumor was considered a potential histologic diagnosis in one of the patients in our study who had a mass that had increased in size at follow-up, but the diagnosis was changed to benign fibroadenoma. Both fibroadenoma and phyllodes tumor manifest as lobular, oval, or round masses. Compared with fibroadenoma, however,

phyllodes tumor more commonly has ultrasound features of ill-defined margins with small cystic spaces [16]. In our study, the only malignant tumor diagnosed was a very small, round, circumscribed hypoechoic mass seen only with ultrasound. Because of the small size, it is unclear how well the imaging finding corresponded to the palpable clinical finding. A very small focus of DCIS (1.5 mm) also was found in this area. It is difficult to know whether the clinical finding was the small hypoechoic mass or the DCIS was an incidental finding. Nevertheless, this finding yields a cancer incidence for palpable complicated cyst versus solid mass of 1.5% (one of 66 masses). Our findings are comparable with those of Berg et al. [17], who found no cancer among 38 nonpalpable complicated cysts. The results of our study are similar to those of Graf et al. [11], who found no cancer in follow-up of 156 palpable lesions with benign mammographic and sonographic features. The results of our study also are similar to those of the original work by Sickles [2], who evaluated nonpalpable lesions that had a benign appearance. On the basis of the results of our study, we suggest that the recommendation for BI-RADS category 3 assessments, initial short-term follow-up (usually 6-month) examination, is a reasonable alternative to biopsy of palpable lesions with benign imaging features and nonsuspicious findings during a focused clinical breast examination. This recommendation may be particularly useful in the care of young patients with probable fibroadenoma, who were the majority of patients in our study. Focused clinical examination and correlation with clinical history are important features in the evaluation of palpable lesions [18]. All patients in this study underwent focused clinical examination of the palpable finding by an attending radiologist. Cases in which the findings at focused breast examination were clinically suspicious for breast cancer, such as the presence of a hard mass or induration, were not recommended for follow-up but for immediate evaluation. Clinical history can be helpful, especially if the patient performs breast self-examination regularly and finds a new palpable lump. However, clinical history also can be misleading owing to denial and other psychosocial issues. Comparison with older images, if available, is important in evaluation of palpable lesions. If a solid mass, palpable or not, is new at imaging in comparison with

AJR:193, December 2009

Follow-Up Imaging of Palpable Breast Lesions old images, biopsy is needed unless a specific benign diagnosis, such as lymph node or oil cyst, can be made. All of the ultrasound examinations in our study were performed by an attending radiologist, who paid close attention to the ultrasound features of the lesion at real-time examination. Two issues cause concern among radiologists about performing short-term followup rather than biopsy of a palpable finding. First, many radiologists may not be confident in their clinical examination skills and may prefer to base recommendations solely on imaging features. However, many radiologists who perform breast imaging have become primary caregivers in breast health, providing services, such as biopsy, counseling, and clinical examination, that may have been provided by surgeons in the past. In our practice, most women with clinical breast findings, including breast pain, palpable masses, and nipple discharge, undergo evaluation by radiologists. Few patients are referred to surgeons by their health care providers when imaging findings are negative or benign. This practice is likely an evolution of breast care originating with the use of percutaneous biopsy in the early 1990s. Thus the use of short-term follow-up when a lesion is palpable will likely rely on radiologists’ comfort with clinical evaluation. The second concern of radiologists is liability for the outcome of a clinical finding. The fourth edition of BI-RADS [1] states only that “short-term follow up of a palpable lesion is not justified by scientific studies.” The results of the study by Graf et al. [11] and our study begin to justify the use of a probably benign assessment in the setting of a palpable finding. In addition, a nonpalpable lesion can be monitored only with imaging. A palpable lesion can be monitored with both imaging and examination. All patients are instructed to return promptly if they or their health care providers notice a change in examination findings. Repetition of the clinical examination by the referring health care provider in 2–3 months also may reassure both patient and physician that the lesion is likely benign. Our study had limitations. It was a retrospective nonrandomized study with inherent selection bias. We excluded palpable lesions for which short-term follow-up was recommended but had descriptors that were not benign, such as irregular shape or ill-defined margins. Inclusion of these lesions would likely have increased the number of cases

AJR:193, December 2009

of cancer found at follow-up because lesions to which short-term follow-up criteria were misapplied would have been included. Rosen et al. [19], in a retrospective study, found that most cases of cancer detected after the recommendation for short-term follow-up of nonpalpable lesions were cases in which criteria had been misapplied. If other radiologists choose to perform short-term follow-up of palpable masses, we highly recommend adherence to this strategy for lesions with benign imaging characteristics, including round, oval, or lobular shape; circumscribed margins; homogeneous ultrasound echotexture; and the absence of malignant characteristics such as posterior acoustic shadowing or echogenic halo at ultrasound examination. In addition, if the mass represents a new finding compared with the findings on older mammograms, biopsy rather than follow-up should be performed. In other words, palpable masses for which short-term follow-up is considered should meet all of the same imaging criteria for short-term follow-up as nonpalpable lesions do. Other limitations of our study were the 12-month rather than 24-month follow-up period and the use of clinical as well as imaging assessment of stability. In the evaluation of short-term follow-up of nonpalpable probably benign lesions, Sickles [2] found that 10 of 15 cases of cancer were diagnosed on the basis of progression in the first 12 months. The use of 12 rather than 24 months of stability in our study therefore might have resulted in reduced ability to detect slowergrowing malignant lesions. Although we request that the patient return for imaging at 24 months, there may be less motivation on the part of both patient and physician for her to return for imaging if the mass is stable at clinical examination at 24 months. Likewise, although imaging follow-up would be ideal for all patients, some women decline imaging follow-up if the mass is stable at clinical examination. Because clinical examination is less sensitive to small changes in mass size than imaging is, slow-growing malignant lesions can be missed with this approach. The mean follow-up period, however, was nearly 1 year longer for women who undertook clinical follow-up (3.6 years) than for the entire group (mean, 2.7 years). A final limitation was that our state cancer registry was not a Surveillance Epidemiology and End Results site. Therefore, we did not have follow-up data on the 13.9% of patients who left our institution. This percent-

age is higher than that in studies of nonpalpable lesions in the screening population among whom lack of compliance is reported as 5.1% [20]. This finding may have been be due to the young age of the patients in our study (mean age, 38.7 years, compared with 52.6 years in the study by Sickles [2]). Younger women may be more transient in our community and may not have health insurance. Patients also may not return if they believe their breast self-examinations show the palpable finding is not changing or has resolved. The results of our study show that because of an acceptably low incidence of breast cancer, short-interval follow-up of palpable masses with benign imaging features is a reasonable alternative to immediate biopsy. Most of the masses followed up in this study were thought prospectively to represent fibroadenoma in young women. Most of the women in our study chose imaging follow-up over immediate biopsy. The cost of follow-up of these palpable masses that have benign features is lower than that of immediate biopsy, even when subsequent biopsy due to interval growth of 7% of lesions is taken into account. Application of the results of our study may reduce the number of biopsies that result in benign findings. Additional studies confirming our findings would be useful. Evaluation by other investigators of the outcome of BIRADS category 4A masses that have benign features but are biopsied because they are palpable at examination would provide additional evidence supporting follow-up of palpable masses that have benign features. References 1. American College of Radiology. Breast imaging reporting and data system (BI-RADS), 4th ed. Reston, VA: American College of Radiology, 2003 2. Sickles EA. Periodic mammographic follow-up of probably benign lesions: results in 3184 consecutive cases. Radiology 1991; 179:463–468 3. Varas X, Leborgne F, Leborgne JH. Nonpalpable, probably benign lesions: role of follow-up mammography. Radiology 1992; 184:409–414 4. Sickles EA. Nonpalpable, circumscribed, noncalcified solid breast masses: likelihood of malignancy based on lesion size and age of patient. Radiology 1994; 192:439–442 5. Sickles EA. Probably benign breast lesions: when should follow-up be recommended and what is the optimal follow-up protocol? Radiology 1999; 213: 11–14 6. American College of Radiology. Breast imaging reporting and data system (BI-RADS), 3rd ed. Reston, VA: American College of Radiology, 1998

1729

Harvey et al. 7. Stavros AT, Thickman D, Rapp CL, Dennis MA, Parker SH, Sisney GA. Solid breast nodules: use of sonography to distinguish between benign and malignant lesions. Radiology 1995; 196:123–134 8. Graf O, Helbich TH, Hopf G, Graf C, Sickles EA. Probably benign breast masses at US: is follow-up an acceptable alternative to biopsy? Radiology 2007; 244:87–93 9. Kerlikowske K, Smith-Bindman R, Abraham LA, et al. Breast cancer yield for screening mammographic examinations with recommendation for short-interval follow-up. Radiology 2005; 234: 684–692 10. Vizcaino I, Gadea L, Andreo L, et al. Short-term follow-up results in 795 nonpalpable probably benign lesions detected at screening mammography. Radiology 2001; 219:475–483 11. Graf O, Helbich TH, Fuchsjaeger MH, et al. Follow-

up of palpable circumscribed noncalcified solid breast masses at mammography and US: can biopsy be averted? Radiology 2004; 233:850–856 12. Sumkin JH, Perrone AM, Harris KM, Nath ME, Amortegui AJ, Weinstein BJ. Lactating adenoma: US features and literature review. Radiology 1998; 206:271–274 13. Santen RJ, Cerilli LA, Harvey JA. The breast: gynecomastia and benign breast disease in women. In: Besser GM, Thorner MO, eds. Comprehensive clinical endocrinology. London, UK: Elsevier Science, 2002:585–612 14. Skaane P, Engedal K. Analysis of sonographic features in the differentiation of fibroadenoma and invasive ductal carcinoma. AJR 1998; 170: 109–114 15. Gordon PB, Gagnon FA, Lanzkowsky L. Solid breast masses diagnosed as fibroadenoma at fine-needle

aspiration biopsy: acceptable rates of growth at longterm follow-up. Radiology 2003; 229:233–238 16. Liberman L, Bonaccio E, Hamele-Bena D, Abramson AF, Cohen MA, Dershaw DD. Benign and malignant phyllodes tumors: mammographic and sonographic findings. Radiology 1996; 198:121–124 17. Berg WA, Campassi CI, Ioffe OB. Cystic lesions of the breast: sonographic–pathologic correlation. Radiology 2003; 227:183–191 18. Harvey JA. Sonography of palpable breast masses. Semin Ultrasound CT MR 2006; 27:284–297 19. Rosen EL, Baker JA, Soo MS. Malignant lesions initially subjected to short-term mammographic follow-up. Radiology 2002; 223:221–228 20. Varas X, Leborgne JH, Leborgne F, Mezzera J, Jaumandreu S, Leborgne F. Revisiting the mammographic follow-up of BI-RADS category 3 lesions. AJR 2002; 179:691–695

F O R YO U R I N F O R M AT I O N

Got a few minutes for a visit? Stop by the American Roentgen Ray Society’s online women’s imaging community, which features exclusive content and resources by and for women’s imaging specialists, including articles, electronic exhibits, webcasts, current news, and job and meeting listings. ARRS provides these valuable resources as a free service just for you. Visit us at http://womensimagingonline.arrs.org/.

1730

AJR:193, December 2009