Abstract
Objective:
To identify the predictive factors of cancer invading into the nipple.
Methods:
Patients with breast cancer undergoing mastectomy between May 2009 and March 2019 were reviewed retrospectively. Of these, those with breast cancer within 2 cm of the nipple areolar complex on ultrasonography were included in this study. Clinicopathological data of the primary tumor and imaging findings from mammography, ultrasonography, and MRI were compared between cases with and without nipple involvement by cancer.
Results:
In total, 156 of the 821 patients identified were included in the analysis. Of them, 29 had nipple involvement by cancer. Univariate analysis revealed that the following imaging results were significantly associated with nipple involvement: perineural invasion, lymphovascular invasion, lymph node metastasis; relation type between the tumor and the nipple on ultrasonography; periareolar skin thickening on mammography; and short tumor-nipple distance, continuous enhancement between the nipple and tumor, skin enhancement, and nipple enhancement on MRI. However, on multivariate logistic regression analysis, only invasion type of tumor on ultrasonography and nipple enhancement and short tumor-nipple distance on MRI were significantly correlated with nipple involvement by cancer.
Conclusion:
Imaging findings on preoperative mammography, ultrasonography and MRI are effective predictors for nipple involvement by cancer.
Advances in knowledge:
Preoperative mammography, ultrasonography, and MRI help predict nipple involvement by breast cancer.
Introduction
The conservation of the nipple areolar complex is an important concern psychologically and cosmetically for breast cancer patients who require surgery. Breast reconstruction techniques that spare the nipple have evolved to improve the overall quality of life for females who have undergone surgery for breast cancer.1
The nipple areolar complex is a specialized anatomical and functional area of the breast.2 It serves to drain and express milk during lactation. The breast consists of approximately 15–20 segments demarcated by collecting ducts that coalesce in the subareolar region into lactiferous sinuses of the nipple areolar complex.3 Terminal duct lobular unit, the site from which breast cancer arises, are seen within the nipple in approximately 9–24% of mastectomy specimens.4–7 Conserving the nipple areolar complex has a potential risk of leaving an occult nipple involvement by cancer.8 Multiple studies of mastectomy specimens have shown that the incidence of nipple involvement by cancer ranges from 5.6 to 58%.3,9–14
Several studies have been conducted to predict nipple involvement by cancer through imaging modalities, with MRI and mammography having been proven to be useful tools for the prediction. These studies have revealed that nipple enhancement, short tumor-nipple distance and continuous enhancement between the tumor and the nipple on MRI, and short tumor-nipple distance on mammography are associated with nipple involvement by cancer.15–22
However, studies evaluating the usefulness of ultrasonography for predicting nipple involvement in breast cancer are lacking to date.18Evaluating subareolar tumors via ultrasonography may be more difficult than tumors located elsewhere in the breast due to anatomical complexity, which is potentially caused by the fibrous composition and the convergence of collecting ducts and lactiferous sinuses.3,23 Despite this limitation, ultrasonography is a convenient imaging modality for the diagnosis and evaluation of malignancy and is often more accessible than other imaging modalities. The purpose of this study was to identify preoperative multimodality image findings including ultrasonography and clinicopathological characteristics that correlate with nipple involvement by cancer.
Methods and materials
Patients
This study adhered to the ethical tenets of the Declaration of Helsinki and was approved by the institutional review board. The requirement for patients’ informed consent was waived owing to the retrospective nature of the study.
We included breast cancer patients who underwent mastectomy from May 2009 to March 2019 at Ulsan University Hospital. Patients with a history of neoadjuvant chemotherapy, excision for biopsy, and Paget’s disease without underlying mammary malignancy were excluded. Two radiologists reviewed the medical records and ultrasonographic images of all patients to determine if each patient met the enrollment criteria. All the patients had undergone mastectomy with histological confirmation within 1 month of preoperative imaging evaluation.
Pathology
The sagittal sections of the entire nipple areolar complex were routinely evaluated for all mastectomy cases. After hematoxylin and eosin staining, each section was reviewed by breast pathologists. Additional sections were obtained or immunohistochemical staining was performed as necessary for diagnosis.
We reviewed the patients’ clinicopathological data, including age, body mass index, pathologic tumor size, pathologic type, histologic grade, lymph node status, distant metastasis, extensive intraductal component, lymphovascular invasion, perineural invasion, p53 mutation, hormone receptor status, human epidermal growth factor receptor 2status and Ki-63.
Techniques of imaging studies
Mammography was performed using three different full-field digital machines during the study period (Senograph DS, GE Healthcare, Milwaukee, WI; Brestige, Medifuture, Seongnam-si, South Korea; Selenia Dimensions, Hologic, Bedford, MA). Standard craniocaudal and mediolateral oblique views were routinely acquired, and supplementary mammographic views were examined as needed.
All ultrasonographic examinations were performed by one of three breast radiologists using a handheld 7–15-MHz linear-array transducer (iU22 Ultra-sound system; Philips Ultrasound, Bothell, WA) at Ulsan University Hospital. These radiologists have 25, 20, and 5 years of experience in breast imaging.
MRI was performed using a 1.5 T system (Achieva; Philips, Best, The Netherlands) and two 3.0 T systems (Achieva; Philips, Best, The Netherlands, MAGNETOM Skyra; Siemens, Erlangen). A breast coil was available on each system. All patients were examined in a prone position. The standard MRI protocol consists of the following sequences: an axial T2 weighted sequence; diffusion-weighted sequence; an unenhanced and contrast-enhanced axial T1 weighted sequence with fat saturation. Six dynamic sequences were performed before and after intravenous injection of the contrast medium [0.1 mmol kg−1 gadoterate dimeglumine (Dotarem, Guerbet, Villepinte, France)]. Post-processing images included production of 1 or 2 min-subtraction, reverse subtraction, maximum intensity projection images and multiplanar reconstruction in the sagittal plane.
Image interpretation
Mammography, ultrasonography, and MRI findings were retrospectively reviewed by two radiologists with 20 years and 5 years of experience in breast imaging who were blinded to the pathological report. Consensus for disagreements on categorization was obtained through a discussion.
Mammographic images were retrospectively reviewed. We analyzed mammographic density according to the Breast Imaging Reporting and Data System (BI-RADS) lexicon. Any nipple changes such as retraction or inversion, periareolar skin thickening, microcalcifications in the nipple and within 0.5 cm from the nipple, mammographic feature such as calcification, without calcification, and undetected and distribution of calcifications were presented. Periareolar skin thickening was considered to be positive if it is thicker by 2 mm compared with the contralateral nipple areolar complex on both craniocaudal and mediolateral oblique views.
The relationship between tumor and nipple areolar complex was assessed using ultrasonography. All tumors were categorized into three types based on this imaging data: (1) subareolar type: tumors located below the nipple but not contacting the nipple; (2) duct dilatation type: tumors located below the nipple with duct dilatation; and (3) invasion type: tumors that contacted or invaded any part of the nipple (Figure 1). We also analyzed background parenchymal echogenicity according to the BI-RADS lexicon, tumor presentation type on ultrasonography such as mass or non-mass, and presence of calcifications on ultrasonography.
Figure 1. .
Three types of relationship between the tumor and the nipple areolar complex on ultrasonography. (a) Subareolar type: tumors located below the nipple but not contacting the nipple. (b) Duct dilatation type: tumors located below the nipple with duct dilatation. (c) Invasion type: tumors that contacted or invaded any part of the nipple areolar complex.
MRI was used to evaluate the continuous enhancement between the tumor and the nipple, tumor–nipple distance, the largest dimension of each enhancing tumor, tumor presentation type such as mass or non-mass, and skin enhancement on 1 or 2 min subtraction sequence. The enhancement patterns of the nipple areolar complex were analyzed and categorized as linear, focal, diffuse heterogeneous, or diffuse homogeneous. Nipple enhancement was also analyzed and categorized into non-enhancement, unilateral enhancement, and bilateral enhancement.
Statistical analysis
All statistical analyses were performed using SPSS statistical software(v. 21.0; IBM Corp., Armonk, NY) and a p-value of less than 0.05 was considered statistically significant. We used Fisher’s exact test, Pearson’s χ2 test, or an independent t-test to evaluate the associations between nipple involvement by cancer and clinicopathological factors or imaging findings. The parameters found to be significant in univariate analysis were analyzed via multivariate analysis using logistic regression with forward covariate selection (Wald test).
Results
In total, 170 patients with subareolar mass or non-mass lesion within 2 cm of the nipple on ultrasonography were included. Of these, 14 were excluded because they did not have available MR images (n = 9), have uncertain pathological results for the nipple areolar complex (n = 4), and did not undergo pre-operative mammography (n = 1). Finally, 156 patients were included in the analysis. Nipple involvement by cancer was found in 29 of the 156 mastectomy specimens (18.5%). 22 patients had invasive ductal carcinoma, 4 had Paget’s disease, 2 had ductal carcinoma in situ, and 1 had mucinous carcinoma. The four patients with Paget’s disease had underlying breast cancer (three invasive ductal carcinoma and one invasive micropapillary carcinoma).
The clinicopathological characteristics of the cases in this study are summarized in Table 1. Lymph node metastasis, lymphovascular invasion, and perineural invasion were significantly different between the pathology-proven nipple involvement group and the pathology-proven non-involvement group.
Table 1. .
Clinicopathological factors of patients according to nipple areolar complex involvement by cancer
| Pathologic diagnosis of nipple areolar complex | p-value | Sensitivity | Specificity | |||
| Negative | Positive | |||||
| Mean | Mean | |||||
| Age | 54.3 | 54.4 | 0.977 | |||
| Body mass index | 24.0 | 24.4 | 0.611 | |||
| Pathologic size | 3.5 | 3.7 | 0.730 | |||
| N | N | |||||
| Site | Right | 60 | 13 | 0.814 | ||
| Left | 67 | 16 | ||||
| Pathology | DCIS | 23 | 2 | 0.445 | ||
| IDC | 94 | 25 | ||||
| ILC | 4 | 1 | ||||
| Mucinous cancer | 3 | 0 | ||||
| Invasive micropapillary cancer | 1 | 1 | ||||
| Papillary carcinoma | 2 | 0 | ||||
| Multifocality | Absent | 119 | 25 | 0.172 | ||
| Present | 8 | 4 | ||||
| Multicentricity | Absent | 111 | 25 | 0.862 | ||
| Present | 16 | 4 | ||||
| Metastasis | Absent | 124 | 29 | 1.000 | ||
| Present | 3 | 0 | ||||
| Lymph node metastasis | Absent | 78 | 10 | 0.008 | 65.5% | 61.4% |
| Present | 49 | 19 | ||||
| Lymph node stage | 0 | 78 | 10 | 0.067 | ||
| 1 | 35 | 13 | ||||
| 2 | 8 | 3 | ||||
| 3 | 6 | 3 | ||||
| Histologic grade(n = 128) | 1 | 15 | 3 | 0.217 | ||
| 2 | 54 | 9 | ||||
| 3 | 33 | 14 | ||||
| Extensive intraductal component(n = 131) | Absent | 69 | 19 | 0.692 | ||
| Present | 35 | 8 | ||||
| Lymphovascular invasion(n = 153) | Absent | 83 | 12 | 0.020 | 57.1% | 66.4% |
| Present | 42 | 16 | ||||
| Perineural invasion(n = 153) | Absent | 109 | 18 | 0.004 | 35.7% | 87.2% |
| Present | 16 | 10 | ||||
| P53 mutation | Absent | 85 | 18 | 0.618 | ||
| Present | 42 | 11 | ||||
| Estrogen Receptor | Negative | 30 | 8 | 0.654 | ||
| Positive | 97 | 21 | ||||
| Progesterone Receptor | Negative | 38 | 12 | 0.233 | ||
| Positive | 89 | 17 | ||||
| Her 2 | Negative | 44 | 13 | 0.304 | ||
| Positive | 83 | 16 | ||||
| Ki-67(n = 125) | Negative | 36 | 9 | 0.865 | ||
| Positive | 65 | 15 | ||||
DCIS, ductal carcinoma in situ; IDC, invasive ductal carcinoma; ILC, invasive lobular carcinoma; Her2, human epidermal growth factor receptor 2.
Mammographic findings
Periareolar skin thickening was significantly different between the pathology-proven nipple involvement group and the pathology-proven non-involvement group. In the pathology-proven nipple involvement group, 20 of 29 cases (69%) showed positive periareolar skin thickening, whereas 25 of 129 cases (19.7%) in the pathology-proven non-involvement group showed positive periareolar skin thickening (p = 0.000). Meanwhile, nipple change, suspicious calcifications in nipple, tumor presentation type, mammographic density, and distribution of calcification were not significantly different between the two groups (Table 2).
Table 2. .
Imaging findings of patients according to nipple areolar complex involvement by cancer
| Pathologic diagnosis of nipple areolar complex | P-value | Sensitivity | Specificity | |||
| Negative | Positive | |||||
| Ultrasonography | ||||||
| Relation type between the tumor and the nipple | Subareolar type | 48 | 4 | 0.000 | 13.7% | 62.2% |
| Duct dilatation type | 42 | 3 | 10.3% | 66.9% | ||
| Invasion type | 37 | 22 | 75.9% | 70.8% | ||
| Tumor presentation type | Mass | 69 | 16 | 0.935 | ||
| Non-mass | 58 | 13 | ||||
| Calcification on US | Absent | 88 | 17 | 0.269 | ||
| Present | 39 | 12 | ||||
| Background parenchymal echogenicity | Homogeneous-fat | 19 | 5 | 0.425 | ||
| Homogeneous-fibroglandular | 101 | 24 | ||||
| Heterogeneous | 7 | 0 | ||||
| Mammography | ||||||
| Nipple retraction/inversion | Absent | 80 | 13 | 0.072 | ||
| Present | 47 | 16 | ||||
| Periareolar thickening | Absent | 102 | 11 | 0.000 | 62.0% | 80.3% |
| Present | 25 | 18 | ||||
| Calcification in nipple | Absent | 94 | 20 | 0.580 | ||
| Present | 33 | 9 | ||||
| Tumor presentation type | Calcification | 65 | 19 | 0.253 | ||
| W/o calcification | 47 | 9 | ||||
| Undetected | 15 | 1 | ||||
| Density | A | 20 | 1 | 0.123 | ||
| B | 35 | 11 | ||||
| C | 41 | 13 | ||||
| D | 31 | 4 | ||||
| Distribution of calcification(N = 87) | Diffuse | 5 | 3 | 0.099 | ||
| Segmental | 38 | 8 | ||||
| Grouped | 21 | 5 | ||||
| Linear | 0 | 1 | ||||
| Multifocal | 3 | 3 | ||||
| MRI | ||||||
| Continuity | Absent | 93 | 4 | 0.000 | 86.2% | 73.2% |
| Present | 34 | 25 | ||||
| Enhancing pattern of the nipple | Absent | 68 | 1 | 0.000 | 3.4% | 46.4% |
| Linear | 21 | 6 | 20.6% | 83.4% | ||
| Focal | 12 | 7 | 24.1% | 90.5% | ||
| Diffuse heterogeneous | 22 | 4 | 13.7% | 82.6% | ||
| Diffuse homogeneous | 4 | 11 | 37.9% | 97.6% | ||
| Tumor presentation type | Mass | 72 | 15 | 0.627 | ||
| Non-mass | 55 | 14 | ||||
| Skin enhancement | Absent | 114 | 20 | 0.004 | 31.0% | 89.7% |
| Present | 13 | 9 | ||||
| Nipple enhancement | Non-enhancement | 74 | 1 | 0.000 | 3.4% | 41.7% |
| Unilateral | 30 | 25 | 86.2% | 76.3% | ||
| Bilateral | 23 | 3 | 17.2% | 81.8% | ||
| Mean(SD) | Mean(SD) | |||||
| Enhancing lesion size | 3.4 (2.0) | 3.9 (2.2) | 0.213 | |||
| Tumor-nipple distance | 1.1 (1.1) | 0.2 (0.5) | 0.000 | |||
SD, standard deviation.
Ultrasonographic findings
The relationship between the tumor and the nipple areolar complex on ultrasonography was significantly different between the two groups (p = 0.000). In total, of the 29 lesions in the pathology-proven nipple involvement group, 22 (75.9%),4 (13.8%), and 3 (10.3%) were categorized as invasion type, subareolar type, and duct dilatation type, respectively. Meanwhile, tumor presentation type, presence of calcifications on ultrasonography, and background parenchymal echogenicity pattern were not significantly different between two groups (Table 2).
MRI findings
Continuous enhancement between the tumor and the nipple was seen in 59 lesions, with statistically significant nipple involvement (p = 0.000). In the pathology-proven nipple involvement group, 25 of 29 lesions (86.2%) showed continuous enhancement between the tumor and the nipple, while 34 (26.8%) of 127 lesions in the pathology proven non-involvement group showed continuous enhancement (p = 0.000).
Skin enhancement was seen in 9/29 lesions (31%) and 13/127 lesions (11.2%) in the pathology-proven nipple involvement group and the pathology-proven non-involvement group, respectively (p = 0.004).
The pattern of nipple enhancement was significantly different between the pathology-proven nipple involvement group and the pathology-proven non-involvement group (p = 0.000). In the pathology-proven involvement group, 11 (37.9%), 7 (24.1%), 6 (20.7%), 4 (13.8%), and 1 (3.4%) of the 29 cases showed diffuse homogeneous, focal, linear, diffuse heterogeneous, and non-enhancement patterns, respectively.
The nipple enhancement was significant statistically (p = 0.000). The nipples with pathology-proven nipple involvement group showed 25 unilateral enhancement (86.2%), 3 bilateral enhancement (10.3%) and 1 non-enhancement (3.4%).
The mean tumor-nipple distance on MRI was significantly different between the pathology-proven nipple involvement group and the pathology-proven non-involvement group (0.2 vs 1.1 cm).
Meanwhile, the tumor presentation type and size of enhancing lesion on MRI were not significantly different between the two groups (Table 2).
In multivariate logistic regression analysis of the three imaging modalities, invasion type of the relation type between the tumor and nipple on ultrasonography [odds ratio, 4.635; 95% confidence interval (CI) (1.034–20.775); p = 0.045], short tumor-nipple distance on MRI [odds ratio, 0.208; 95% CI (0.074–0.582); p = 0.003], unilateral [odds ratio, 75.477; 95% CI (8.567–664.945); p = 0.000] and bilateral [odds ratio, 12.131; 95% CI (1.083–135.934); p = 0.043] enhancement of the nipple on MRI remained independent and statistically significant predictors of nipple involvement (Figure 2, Table 3).
Figure 2. .
A 64-year-old female with invasive ductal carcinoma in the left breast. (a) Ultrasonographic image shows breast cancer (arrows) invading the nipple. (b) Contrast-enhanced T1 weighted MR image shows breast cancer with unilateral diffuse homogeneous enhancing nipple (arrow) in the left breast.
Table 3. .
Multivariate analysis of clinicopathologic factors and imagingfindings
| Bias(SE) | OR | 95% CI | P-value | |
| Ultrasonography | ||||
| Tumor type based on US imaging (Relationship between the tumor and the nipple) | ||||
| Subareolar type | Reference | |||
| Duct dilatation type | −1.070 (0.922) | 0.343 | 0.056–2.089 | 0.246 |
| Invasion type | 1.526 (0.819) | 4.635 | 1.034–20.775 | 0.045 |
| MRI | ||||
| Nipple enhancement | ||||
| Non-enhancement | Reference | |||
| Unilateral | 4.324 (1.100) | 75.477 | 8.567–664.945 | 0.000 |
| Bilateral | 2.496 (1.233) | 12.131 | 1.083–135.934 | 0.043 |
| Tumor-nipple distance | ||||
| −1.572 (0.526) | 0.208 | 0.074–0.582 | 0.003 | |
SE, standard error; OR, odds ratio; CI, confidence interval.
The image findings of the pathology-proven proven nipple involvement group are attached to Supplementary Material 1.
Discussion
In this study, we found that ultrasound data might be a significant predictor of nipple involvement by cancer by multivariate logistic regression analyses. Ultrasound data were analyzed in a previously published report by Hwang.18 In this report, the margin and orientation of the primary tumor were statistically significantly associated with nipple involvement according to Fisher’s exact test, but not in the multivariate analysis. We did not analyze the characteristics of primary tumor on ultrsonography according to the BI-RADS lexicon, but we focused on the relationship between the tumor and the nipple and categorized to three types such as subareolar, duct dilatation and invasion type. In this study, invasion type on ultrasonography has a significant higher nipple involvement rate than subareolar type and duct dilatation type.
Unilateral and bilateral enhancement of the nipple areolar complex on MRI was significantly associated with nipple involvement by cancer in both univariate and multivariate logistic regression analyses. Although bilateral enhancement of the nipple was more significantly correlated with nipple invasion than non-enhancement, unilateral enhancement of the nipple was a stronger and more significant predictor of the nipple involvement than bilateral enhancement (odds ratio 75.477 vs 12.131) on multivariate logistic regression analysis. Further, short tumor–nipple distance, continuous enhancement between the tumor and the nipple, enhancing pattern of the nipple, and skin enhancement were also associated with nipple involvement by cancer. These results are consistent with previous studies15–17showing that unilaterally enhanced nipple with continuous enhancement, nipple areolar complex enhancement, and short tumor–nipple distance were predictive factors for nipple involvement by cancer.
Majority of cases of subareolar type and duct dilatation type were not associated with nipple involvement by cancer. However, four cases of subareolar type and three cases of duct dilation type on ultrasonography showed pathology-proven nipple involvement by cancer. This forms the basis for concerns of surgeons about nipple preservation because the nipple and adjacent ducts may contain cancer cells that have spread distally along the ducts from the primary cancer.24 Continuity between the tumor and the nipple was not seen on ultrasonography for these seven cases, but unilateral enhancement of the nipple on MRI was seen in all seven cases and linear enhancement was seen in six cases (Figures 3 and 4). Thus, the complementary use of ultrasonography and MRI helps to increase predictability of nipple involvement by breast cancer.
Figure 3. .
A 26-year-old female with invasive ductal carcinoma in the left breast. (a) Ultrasonographic image shows non-mass lesion (arrow) without directly connection with the nipple in subareolar area. (b) Contrast-enhanced T1 weighted MR image shows linear enhancement (arrow) in the nipple. (c) Mammography shows multifocal clustered microcalcifications (arrows), but no other findings in the left breast. (d) Histologic nipple tissue specimen shows invasive ductal carcinoma involving clusters of several ducts of the nipple (arrows). (Hematoxylin-eosin stain; original magnification, (X20).)
Figure 4. .
A 45-year-old female with invasive ductal carcinoma in the right breast. (a) Ultrasonographic image shows a circumscribed and round-shaped mass (solid arrow) with duct dilatation (open arrow). (b) Contrast-enhanced T1 weighted MR image shows a linear enhancement (arrow) in the nipple. (c) Mammography shows asymmetry with suspicious microcalcifications (open arrow) associated with periareolar skin thickening and nipple retraction(solid arrow). (d) Histologic nipple tissue specimen shows invasive ductal carcinoma in dilated duct of the nipple (arrow). (Hematoxylin-eosin stain; original magnification, (X20).)
Previous studies have shown that short tumor-nipple distance on mammography is a useful predictor of nipple involvement by cancer,18,22 but this was not measured in our study. Any imaging modality examination of the tumor–nipple distance is expected to produce a similar result.19Because mammographic distance may vary depending on the patient’s clinical symptoms or compression level, the tumor–nipple distance measured via MRI is considered more objective than other imaging modalities. In this study, periareolar skin thickening was a statistically significant predictor of nipple involvement by cancer, although it was only significant in univariate logistic regression analysis.
Of the clinicopathological factors evaluated in this study, perineural invasion, lymphovascular invasion, and lymph node metastasis were significantly correlated with nipple involvement by cancer. The association between nipple involvement by cancer and perineural invasion has not been studied. However, perineural invasion is known to be a poor prognostic factor associated with higher T-stage, tumor grade, and lymphovascular invasion.25 Lymph node metastasis and lymphovascular invasion have been shown to be associated with nipple involvement by cancer in several studies, including that by Byon et al26and Kaplan et al.27 These studies also reported an association between nipple involvement by cancer and both p53 mutation and large tumor size.26,27
Limitations
There are several limitations to our study. First, whereas two radiologists subjectively reviewed ultrasonographic images, ultrasonography is a real-time assessment tool. Therefore, there is a possibility that the relationship between the nipple and the tumor was misinterpreted when reviewing ultrasonography images. To avoid controversy about indefinite relationships between the tumor and the nipple, we tried to establish clear criteria for classifying the relationship between the nipple and the tumor in ultrasonography images. Second, we included Paget’s disease in this study, which may complicate our results. However, most cases of Paget’s disease have underlying breast cancer, and treatment planning does not differ from that for cases of nipple involvement by cancer.28 Third, this study was vulnerable to selection bias owing to the retrospective study design. Finally, the sample size was relatively small. To overcome these limitations, a larger population and with more definite histological analyses is needed.
Conclusion
Imaging modalities, including mammography, ultrasonography, and MRI are effective methods to preoperatively predict nipple involvement by cancer. The tumors contacting or invading the nipple on ultrasonography and nipple enhancement and short tumor–nipple distance on MRI were significantly correlated with nipple involvement by cancer. However, the findings of different imaging modalities should be collectively viewed, rather than focusing on only one image finding of each modality, for complementary judgment.
Footnotes
Acknowledgment: We thank all of the partners and staff who contributed to this study.
Contributor Information
Soyeoun Lim, Email: soyeoun.lim.xr@uuh.ulsan.kr.
Gyeongmin Park, Email: 073815@uuh.ulsan.kr.
Hye-jeong Choi, Email: thanksg@uuh.ulsan.kr.
Woon Jung Kwon, Email: becareful123@uuh.ulsan.kr.
Byeong Seong Kang, Email: kbs1radio@uuh.ulsan.kr.
Minseo Bang, Email: bangms@uuh.ulsan.kr.
REFERENCES
- 1.Didier F, Radice D, Gandini S, Bedolis R, Rotmensz N, Maldifassi A, et al. Does nipple preservation in mastectomy improve satisfaction with cosmetic results, psychological adjustment, body image and sexuality? Breast Cancer Res Treat 2009; 118: 623–33. doi: 10.1007/s10549-008-0238-4 [DOI] [PubMed] [Google Scholar]
- 2.An HY, Kim KS, Yu IK, Kim KW, Kim HH. Image presentation. The nipple-areolar complex: a pictorial review of common and uncommon conditions. J Ultrasound Med 2010; 29: 949–62. [DOI] [PubMed] [Google Scholar]
- 3.Sarica O, Zeybek E, Ozturk E. Evaluation of nipple-areola complex with ultrasonography and magnetic resonance imaging. J Comput Assist Tomogr 2010; 34: 575–86. doi: 10.1097/RCT.0b013e3181d74a88 [DOI] [PubMed] [Google Scholar]
- 4.Kryvenko ON, Yoon JY, Chitale DA, Lee MW. Prevalence of terminal duct lobular units and frequency of neoplastic involvement of the nipple in mastectomy. Arch Pathol Lab Med 2013; 137: 955–60. doi: 10.5858/arpa.2012-0137-OA [DOI] [PubMed] [Google Scholar]
- 5.Reynolds C, Davidson JA, Lindor NM, Glazebrook KN, Jakub JW, Degnim AC, et al. Prophylactic and therapeutic mastectomy in BRCA mutation carriers: can the nipple be preserved? Ann Surg Oncol 2011; 18: 3102–9. doi: 10.1245/s10434-011-1908-8 [DOI] [PubMed] [Google Scholar]
- 6.Brachtel EF, Rusby JE, Michaelson JS, Chen LL, Muzikansky A, Smith BL, et al. Occult nipple involvement in breast cancer: clinicopathologic findings in 316 consecutive mastectomy specimens. J Clin Oncol 2009; 27: 4948–54. doi: 10.1200/JCO.2008.20.8785 [DOI] [PubMed] [Google Scholar]
- 7.Rosen PP, Tench W. Lobules in the nipple. frequency and significance for breast cancer treatment. Pathol Annu 1985; 20 Pt 2(Pt 2): 317–22. [PubMed] [Google Scholar]
- 8.Spear SL, Hannan CM, Willey SC, Cocilovo C, mastectomy N-sparing. Nipple-sparing mastectomy. Plast Reconstr Surg 2009; 123: 1665–73. doi: 10.1097/PRS.0b013e3181a64d94 [DOI] [PubMed] [Google Scholar]
- 9.Friedman EP, Hall-Craggs MA, Mumtaz H, Schneidau A. Breast Mr and the appearance of the normal and abnormal nipple. Clin Radiol 1997; 52: 854–61. doi: 10.1016/S0009-9260(97)80081-5 [DOI] [PubMed] [Google Scholar]
- 10.Menon RS, van Geel AN. Cancer of the breast with nipple involvement. Br J Cancer 1989; 59: 81–4. doi: 10.1038/bjc.1989.15 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Smith J, Payne WS, Carney JA. Involvement of the nipple and areola in carcinoma of the breast. Surg Gynecol Obstet 1976; 143: 546–8. [PubMed] [Google Scholar]
- 12.Laronga C, Kemp B, Johnston D, Robb GL, Singletary SE. The incidence of occult nipple-areola complex involvement in breast cancer patients receiving a skin-sparing mastectomy. Ann Surg Oncol 1999; 6: 609–13. doi: 10.1007/s10434-999-0609-z [DOI] [PubMed] [Google Scholar]
- 13.Chang RYK, Cheung PSY. Nipple preservation in breast cancer associated with nipple discharge. World J Surg 2017; 41: 176–83. doi: 10.1007/s00268-016-3679-7 [DOI] [PubMed] [Google Scholar]
- 14.Simmons RM, Brennan M, Christos P, King V, Osborne M. Analysis of nipple/areolar involvement with mastectomy: can the areola be preserved? Ann Surg Oncol 2002; 9: 165–8. doi: 10.1007/BF02557369 [DOI] [PubMed] [Google Scholar]
- 15.Moon JY, Chang Y-W, Lee EH, Seo DY. Malignant invasion of the nipple-areolar complex of the breast: usefulness of breast MRI. AJR Am J Roentgenol 2013; 201: 448–55. doi: 10.2214/AJR.12.9186 [DOI] [PubMed] [Google Scholar]
- 16.Sakamoto N, Tozaki M, Hoshi K, Fukuma E. Is MRI useful for the prediction of nipple involvement? Breast Cancer 2013; 20: 316–22. doi: 10.1007/s12282-012-0338-1 [DOI] [PubMed] [Google Scholar]
- 17.Piato JRM, de Andrade RDJA, Chala LF, de Barros N, Mano MS, Melitto AS, et al. MRI to predict nipple involvement in breast cancer patients. AJR Am J Roentgenol 2016; 206: 1124–30. doi: 10.2214/AJR.15.15187 [DOI] [PubMed] [Google Scholar]
- 18.Hwang H, Park S, Koo JS, Park HS, Kim SI, Cho YU, et al. Factors predictive of occult nipple-areolar complex involvement in patients with carcinoma in situ of the breast. J Surg Oncol 2017; 116: 1046–55. doi: 10.1002/jso.24768 [DOI] [PubMed] [Google Scholar]
- 19.Mariscotti G, Durando M, Houssami N, Berzovini CM, Esposito F, Fasciano M, et al. Preoperative MRI evaluation of lesion-nipple distance in breast cancer patients: thresholds for predicting occult nipple-areola complex involvement. Clin Radiol 2018; 73: 735–43. doi: 10.1016/j.crad.2018.03.008 [DOI] [PubMed] [Google Scholar]
- 20.Cho J, Chung J, Cha E-S, Lee JE, Kim JH. Can preoperative 3-T MRI predict nipple-areolar complex involvement in patients with breast cancer? Clin Imaging 2016; 40: 119–24. doi: 10.1016/j.clinimag.2015.08.002 [DOI] [PubMed] [Google Scholar]
- 21.Machida Y, Shimauchi A, Igarashi T, Hoshi K, Fukuma E. Preoperative breast MRI: reproducibility and significance of findings relevant to nipple-areolar complex involvement. Breast Cancer 2018; 25: 456–63. doi: 10.1007/s12282-018-0845-9 [DOI] [PubMed] [Google Scholar]
- 22.Loewen MJ, Jennings JA, Sherman SR, Slaikeu J, Ebrom PA, Davis AT, et al. Mammographic distance as a predictor of nipple-areola complex involvement in breast cancer. Am J Surg 2008; 195: 391–5. discussion 4-5. doi: 10.1016/j.amjsurg.2007.12.020 [DOI] [PubMed] [Google Scholar]
- 23.Da Costa D, Taddese A, Cure ML, Gerson D, Poppiti R, Esserman LE. Common and unusual diseases of the nipple-areolar complex. Radiographics 2007; 27(Suppl 1): S65–S77. doi: 10.1148/rg.27si075512 [DOI] [PubMed] [Google Scholar]
- 24.Patani N, Devalia H, Anderson A, Mokbel K. Oncological safety and patient satisfaction with skin-sparing mastectomy and immediate breast reconstruction. Surg Oncol 2008; 17: 97–105. doi: 10.1016/j.suronc.2007.11.004 [DOI] [PubMed] [Google Scholar]
- 25.Karak SG, Quatrano N, Buckley J, Ricci A. Prevalence and significance of perineural invasion in invasive breast carcinoma. Conn Med 2010; 74: 17–21. [PubMed] [Google Scholar]
- 26.Byon W, Kim E, Kwon J, Park YL, Park C. Magnetic resonance imaging and clinicopathological factors for the detection of occult nipple involvement in breast cancer patients. J Breast Cancer 2014; 17: 386–92. doi: 10.4048/jbc.2014.17.4.386 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Eisenberg REK, Chan JSY, Swistel AJ, Hoda SA. Pathological evaluation of nipple-sparing mastectomies with emphasis on occult nipple involvement: the Weill-Cornell experience with 325 cases. Breast J 2014; 20: 15–21. doi: 10.1111/tbj.12199 [DOI] [PubMed] [Google Scholar]
- 28.Ashikari R, Park K, Huvos AG, Urban JA. Paget's disease of the breast. Cancer 1970; 26: 680–5. doi: [DOI] [PubMed] [Google Scholar]