Abstract
Background
Breast ultrasound is the major imaging modality in young women because it incurs no radiation exposure and dense breast tissue, which is common in young women, yields a high rate of false-negative results on mammography.
Purpose
To investigate the cancer rates of the sonographic BI-RADS categorization and histopathologic results according to the presence of symptoms in young Asian women.
Material and Methods
We included 811 young women under 30 years of age who underwent a breast ultrasound during the study period. The mean age of all subjects was 24.5 years (range 11–29 years). Histopathologic results were compared with the results after application of the BI-RADS categorization.
Results
Sonographic findings were classified as category 1 (n = 192), category 2 (n = 81), category 3 (n = 399), category 4 (n = 134), and category 5 (n = 5). The cancer rates for category 1, 2, 3, 4, and 5 were 0, 0, 0.3, 6.3, and 100% in the symptomatic group, respectively. For the asymptomatic group, the cancer rates were 0, 0, 0, 8.7, and 100%, respectively. More cancers were found in high-risk women (22.2%, 4/18) than in non-high-risk women (1.4%, 11/793).
Conclusion
BI-RADS categorization was helpful for predicting the probability of malignancy in young Asian women.
Keywords
Young women often complain of palpable breast masses, and most of them are benign masses (1, 2), or prominent breast tissue affected by the stages of the menstrual cycle (3). They are usually referred to a breast examination due to palpable lesions. Recently, however, breast examinations for screening women without symptoms are becoming more common than ever as the awareness of breast diseases increase (1, 4, 5). In addition, the incidence of breast cancer in young women in Korea has increased 154.3% between 1996 and 2004 (5).
Breast ultrasound (US) is the major imaging modality in young women because it incurs no radiation exposure and dense breast tissue, which is common in young women, yields a high rate of false-negative results on mammography (6–9). However, as breast cancer is more common in women over the age of 40, the study of breast diseases has largely been limited to this age group. Until now, the incidence of breast cancer has been reported as lower in young women than that of the older age group, nevertheless, breast cancer in young women tends to be larger and more aggressive and more advanced with metastatic lymph nodes (10–13). Breast cancers in younger women most commonly present as palpable masses but in older women, most breast cancers are detected during screening examinations (11, 14).
However, there have not been many analyses of breast disease or examinations in young aged women. In this study, we intended to investigate the cancer rate of the sonographic BI-RADS categorization and histopathologic results in young Asian women.
Material and Methods
The institutional review board at our institution approved this retrospective study and the requirement for informed consent from the patients was waived. Between January 2005 and December 2006, breast US examinations were performed on 12,380 women. Of these women, 930 (7.5%) were under 30 years of age. Breast US was, therefore, the first choice of breast imaging. These young women underwent a total of 1882 breast US examinations, with each patient having an initial and a follow-up US examination. Of them (n = 930), 110 women who were examined as a follow-up US with 347 examinations were excluded. Of the remaining 820 women, nine had been diagnosed with breast cancer at other clinics and were excluded from this study. A total of 811 young Korean women made up our final study subjects with a mean age of 24.5 years (range 11–29 years).
Breast US examinations were performed by five dedicated breast imaging radiologists, with two to 10 years of experience. The US examinations were performed with a 5–12-MHz linear-phased-array transducer and a commercially available scanner (Antares, Siemens Medical Solutions, Mountain View, CA, USA, ATL HDI 5000 or iU, Philips Medical Systems, Bothell, WA, USA). Mammography was performed when microcalcifications were suspected on breast US or when a lesion was proven malignant histologically with full-field digital mammography (Selenia Full Field Digital Mammography System, Lorad/Hologic, Danbury, CT, USA). Final assessments of the US findings were classified as category 1–5 according to the standard of the American College of Radiology, Breast Imaging Reporting and Data System (BI-RADS). When the lesion was a simple cyst or an intramammary lymph node, or showed hyperechogenicity on US, the lesion was classified as a category 2 (15, 16). A routine annual follow-up examination was recommended for patients with lesions classified as category 1 and 2. When the lesion was a solid mass with an oval or gently lobulated shape, circumscribed margin, and a parallel orientation, it was classified as a category 3 lesion (16–23). Patients with category 3 lesions were followed-up at six month intervals for two years. If the lesion showed one or more of the following suspicious findings: round or irregular shape, microlobulated, indistinct, angular, or spiculated margin, non-parallel orientation, echogenic halo, complex echogenicity, posterior shadowing, duct extension, or microcalcifications within the mass, it was categorized as category 4 or 5 lesion (16, 21). Category 4 was divided into categories 4a, 4b and 4c according to the severity of the suspicious findings. We recommended biopsies for patients with category 4 or 5 breast lesions. Tissue confirmation was also performed if either the patient or clinician wanted to confirm the diagnosis of a breast lesion, or if there was a suspicious clinical scenario despite being classified as category 1, 2, or 3.
Histopathologic results were confirmed via an automated 14-gauge core needle biopsy, fine needle aspiration biopsy (FNAB), vacuum-assisted biopsy, or surgical excision. An US-guided core biopsy was performed using a disposable 14-gauge Tru-cut needle with a 22 mm throw (SACNTM Biopsy Needle, Medical Device Technologies, Gainesville, FL, USA) and five or six core samples were obtained. FNAB with a 23-gauge needle was indicated for complicated and symptomatic cysts. A vacuum-assisted biopsy was performed, using an 11-gauge or 8-gauge probe (Mammotome, Ethicon Endo Surgery, Cincinnati, OH, USA). Vacuum-assisted biopsy was selected for the lesions that had a visible microcalcification on US or discordant pathologic results after core needle biopsy, or the purpose of mass removal. Surgery was performed if preferred by the patient and physician, and in cases of biopsy-proven malignancies. In the absence of histopathologic results, medical records were used to estimate the incidence of breast cancer in the study subjects until June 2009.
Symptoms, past medical history, family history, and the reason for referral were assessed via medical records. The results of the follow-up study were included if patients were followed up more than once. The cancer rates for the five BI-RADS® categories were calculated. We investigated the cancer rates of symptomatic women and asymptomatic women. The cancer rate is defined as the number of breast cancers divided by the total number of examinations per category×100. Fisher's exact test was used for assessing the statistical significance in the cancer rates between symptomatic and asymptomatic women. High-risk patients were defined as patients with a family or personal history of breast cancer, or a history of malignancy other than breast cancer. The cancer rates for high-risk women and non-high-risk women were assessed using a Fisher's exact test and compared with each other. Differences in mean sizes of malignant and benign lesions were assessed for significance using a Wilcoxon rank-sum test. A P value less than 0.05 was considered statistically significant. Statistical analyses were performed using SAS version 9.1 (SAS institute Inc, Cary, NC, USA).
Results
The indications for breast US in the 811 women are listed in Table 1. Of all patients, 72.5% (n = 588) were referred being symptomatic. The most common symptom was a palpable mass (76.7%). In asymptomatic women, the most common reason for a breast US referral was routine screening for breast cancer (n = 111). Screening US were performed for reassurance of patients with family history of breast cancer (n = 8) or fear of breast cancer (n = 92), preliminary examination for hormone replacement therapy for treatment of amenorrhea or Turner syndrome (n = 11). The sonographic BI-RADS final assessment categories and the corresponding management are listed in Table 2.
Indications for breast sonography in 811 women under 30 years of age
*Those with a fear of breast cancer
HRT = hormone replacement therapy
Management according to BI-RADS assessment categories. Numbers in parentheses indicate percentages
Method of follow-up for the final BI-RADS assessment categories
In 306 of the 811 patients, biopsies were performed less than three months after the first US, and follow-up US were performed on 160 patients, but 345 patients failed to undergo a follow-up examination (Table 2). In category 1 and 2, there was a loss in annual routine follow-up in more than 85% in the symptomatic group and 77% in the asymptomatic group, however, in categories 3, 4 and 5, there was less than 23% loss in the symptomatic group and 35% in the asymptomatic group. The 306 biopsies included 14-gauge core needle biopsies (n = 173), vacuum-assisted biopsies (n = 105), excisions (n = 12), fine needle aspirations (n = 15), and a modified radical mastectomy (n = 1). Of the 505 patients without a biopsy, 160 patients underwent a follow-up US at least more than once. The length of time interval between the initial examination and the follow-up examination was as follows (range 6–41 months; median 20 months). One malignancy was found among these 160 patients. Table 3 shows the cancer rates of each BI-RADS category. The cancer rates for categories 1, 2, 3, 4, and 5 were 0, 0, 0.3, 6.3, and 100% in the symptomatic group, respectively. For the asymptomatic group, the cancer rates for the same categories were 0, 0, 0, 8.7, and 100%, respectively. These data include the initial and follow-up US and biopsies (Table 3). There were no significant differences between symptomatic and asymptomatic women for cancer rates.
The cancer rates according to BI-RADS category. Numbers in parentheses indicate the cancer rates
There was no significant difference between symptomatic and asymptomatic women
Final diagnoses proven by histopathologic result
We detected 15 malignancies and 291 benign lesions pathologically (Table 4). The mean size of the malignancies was 22.6 mm (median 22 mm; range 12–34 mm) and that of the benign lesions was 20.4 mm (median 17 mm; range 3–100 mm). There were no significant differences in the mean size of the malignant and benign lesions (p = 0.2183). The most common malignant lesion was invasive ductal carcinoma (Fig. 1). The most common benign lesion was fibroadenoma (Fig. 2). There were one malignant phyllodes tumor that was detected on follow-up US in category 3, invasive ductal carcinoma (n = 5), DCIS (n = 2), metastatic breast malignancy such as neuroendocrine cancer and embryonal rhabdomyosarcoma (n = 2) in category 4 and invasive ductal carcinoma (n = 2), DCIS (n = 2) and acute lymphoblastic leukemia (n = 1) in category 5. Besides, many benign lesions such as fibroadenoma and fibrocystic change were diagnosed. Eleven patients with malignancies were symptomatic (10 patients had palpable lesions; one patient had bloody nipple discharge).
A 27-year-old woman with a palpable mass in her right breast. US showed a 2.5 cm, irregular, hypoechoic mass (arrows). It was classified as BI-RADS category 4 and confirmed as invasive ductal carcinoma by surgery
A 25-year-old woman with a palpable mass in her right breast. US revealed a well-defined oval mass (arrows). It was classified as BI-RADS category 3 and confirmed as fibroadenoma by surgery
Histopathologic results
*Other benign lesions included ductectasia, granulation tissue formation, reactive hyperplasia, hematoma, sclerosing lobular hyperplasia, and fibrosis
Three patients had hypermetabolic lesions on a PET-CT scan without symptoms. They were diagnosed with acute lymphoblastic leukemia (Fig. 3), embryonal rhabdomyosarcoma, and invasive ductal carcinoma. The two patients with acute lymphoblastic leukemia and embryonal rhabdomyosarcoma underwent whole body evaluation via PET-CT scan. Another patient with invasive ductal carcinoma underwent a follow-up PET-CT scan after chemotherapy and this patient was diagnosed with invasive ductal carcinoma in the contralateral breast. All these three patients were well-assigned to category 4 or 5 based on the US results. One patient was referred for a breast US due to the abnormality found in an outside clinic without symptom. The stages of the seven invasive ductal carcinomas were I (n = 1), IIA (n = 3), IIIA (n = 1), IIIB (n = 1), and IIIC (n = 1).
A 20-year-old woman with known acute lymphoblastic leukemia; (a) A PET-CT scan displayed hypermetabolic lesions (arrows) in both breasts; (b) US shows an ill-defined heterogeneously hypoechoic lesion (arrows) in the right breast. It was classified as BI-RADS category 4 and determined to be acute lymphoblastic leukemia via core needle biopsy
Four patients were diagnosed with ductal carcinoma in situ. Two of these patients were diagnosed based on US results only (Fig. 4). The other two patients were diagnosed based on both mammography and US. Follow-up biopsies found a malignant phyllodes tumor in one patient. That lesion had been classified as category 3 on initial US but it rapidly enlarged on the follow-up US, and a biopsy was performed.
A 27-year-old woman with nipple discharge in her right breast. US showed multiple ill-defined tubular hypoechoic lesions (arrows). Ductal carcinoma in situ was confirmed by 8-gauge directional vacuum assisted removal and subsequent mastectomy
Cancer rate in high-risk group
Of the 811 enrolled patients, 18 were considered as a high-risk group. Thirteen of these patients had a positive family history of breast cancer, two had a personal history of breast cancer, and three had a personal history of malignancy other than breast cancer. There were a total of four malignancies among these 18 patients. One had a personal history of breast cancer with invasive ductal carcinoma. The other three malignancies other than breast cancer included acute lymphoblastic leukemia, embryonal rhabdomyosarcoma, and a neuroendocrine tumor. Among the non-high-risk patients (n = 793), there were 11 cases of breast cancer. Six were invasive ductal carcinoma, four were DCIS, and one was a malignant phyllodes tumor. Therefore, the rate of malignancy in the patients with high risk and non-high risk were 22.2% (4/18) and 1.4% (11/793), respectively. This difference was found to be statistically significant (p = 0.0002).
Discussion
Young women often complain of palpable breast lumps, but most of these lumps are benign. The prevalence of primary breast cancer in young women is low, approximately 2% for 20–25 year old women, and 1% in women under 20 years of age (12, 24). Although the incidence of breast cancer is lower in young women than that of the older age group, breast cancer in young women tends to be larger (median tumor size 2 cm in young women compared to 1.5 cm in older women). It also tends to be more aggressive and more advanced with metastatic lymph nodes (10–13). Breast cancers in younger women most commonly present as palpable masses but in older women, most breast cancers are detected during screening mammography (11, 14).
In this study, most US final assessments in younger women fell into BI-RADS category 1 (23.7%) due to prominent breast tissue, or category 3 (49.1%), because of the presence of many benign nodules. The mean size of malignant lesions was 20.6 mm, corresponding to results from prior reports (11, 12). In this study, the cancer rates for category 3 and category 4 were 0.3 and 6.7, respectively. The cancer rates for categories 1, 2, and 5 in the present study were similar to the cancer rates from prior studies (21, 25–28). However, the cancer rate for category 4 was lower than that from previously published studies. The cancer rate for category 4a was reported to be 6% in a previous study (27). The reason for the lower cancer rate in category 4 of this study is assumed to be because most of the category 4 lesions were category 4a (94%, 126/134), rather than category 4b or 4c. That can explain why the total cancer rate of the patients that underwent biopsy (PPV2) was low (4.9%, 15/306) in the present study.
Because high-risk patients had a higher rate of malignancy (22.2%) compared to low-risk patients (1.4%), those who fall into the high-risk group should be closely monitored. Among the high-risk patients, there were four cases of cancer in patients with a personal history of either breast cancer or other malignancies. There were no malignancies found in patients who had a family history of breast cancer. This could be due to the relatively small number of patients who had a family history compared to the entire number of patients in this study. Although none of the patients who were diagnosed with a malignancy had a positive family history, the detection rate of breast cancer in young women who have a family history has been reported to be the same as that of patients older than 50 years of age (29). Therefore, family history of breast cancer can be a predisposing factor that breast exams in young women should be emphasized.
Breast US is performed in elderly patients to detect and localize malignancies and to evaluate non-palpable lesions as an adjunct to mammography (2). It is performed preferentially for young women for both detection of malignancy and evaluation of symptoms. In this study, the most common reason for referral to a breast US was a palpable mass. This corresponds to that of two previous studies that included 600 young women (1, 30). In asymptomatic women, breast US is primarily performed for screening purposes and evaluation of an abnormality found in the course of an examination performed in other hospitals.
US findings fell into category 3 in most patients, regardless of the presence of symptoms. This could be explained by the high frequency of benign lesions in young women, which include fibroadenomas or fibrocystic changes (6). The second most common category was category 1, which is explained by the palpability of normal breast tissue or breast pain due to the menstrual cycle. Most patients were assessed as category 1, 2, or 3 (82.9%) on US. However, only 160 of the patients had a follow-up US without a biopsy more than once and the other patients did not undergo a follow-up US. In these 160 patients, one malignancy was found. This patient had a palpable mass that was rapidly growing and was confirmed as a malignant phyllodes tumor by excision.
Malignancies were found in 15 patients (1.9%, 15/811), including one patient in whom the malignancy was found on a follow-up US. This result (1.9%) is similar to the frequency of malignancy found in young women (1–2%) in previous reports (8, 31). This low frequency of malignancy suggests that the removal of a solid mass in a young woman is not always the best management, as the previous reports supported (3, 8). Of the malignancies, 11 patients had symptoms. Because breast examinations are not routinely performed for those under 40 years of age, the presence of symptoms is a very important indication for breast US in young women. Another three patients without symptoms had hypermetabolic lesions based on a PET-CT scan. They were diagnosed with acute lymphoblastic leukemia, embryonal rhabdomyosarcoma, and invasive ductal carcinoma after subsequent breast US and US-guided core needle biopsy. Although a PET-CT scan has limited ability to detect lesions that are less than 1 cm in diameter or are less aggressive histologic subtypes (32–34), PET-CT scans can help diagnose breast cancer by detecting accelerated metabolic activity that occurs before anatomic structural changes (33).
One limitation of this study is that only 31.7% (160/505) of the patients had more than one follow-up without a biopsy. The remaining 68.3% (345/505) of the patients had no follow-up. This is likely because young women were confident that a negative first US meant that they were disease-free and they failed to follow-up. The management protocol for a category 3 lesion involves a short-term follow-up every six months for two years, and checking the stability of the lesion. However, the management of women under 30 years of age has not been reported, to the best of our knowledge. Therefore, further study with a large study population of young women will be necessary.
In conclusion, BI-RADS categorization was helpful in predicting the probability of malignancy in young Asian women.