British Journal of Radiology (2009) 82, 20-27
© 2009 British Institute of Radiology
doi: 10.1259/bjr/52846080
British Journal of Radiology 82 (2009),20-27 ©2009 The British Institute of Radiology
Do screen-detected lobular and ductal carcinoma present with different mammographic features?
S GARNETT, MSc, BA (Hons), DCR1,
M WALLIS, MBChB, FRCR2 and
G MORGAN, MEd, BA (Hons), TDCR
1 Breast Screening Unit, Ground Floor, West Wing, University Hospital, Clifford Bridge Road, Coventry CV2 2DX, 2 Cambridge Breast Unit, Box 97, Addenbrooke's Hospital, Cambridge CB22 2QQ and 3 Faculty of Health and Social Care Sciences, Kingston University, Penrhyn Road, Kingston upon Thames KT1 2EE, UK
Correspondence: S Garnett, Breast Screening Unit, Ground Floor, West Wing, University Hospital, Clifford Bridge Road, Coventry CV2 2DX, UK. E-mail: sue.garnett{at}uhcw.nhs.uk
 |
Abstract
|
|---|
The aim of this study is to investigate any difference in the shape and location of infiltrating lobular carcinoma (ILC) with respect to the parenchymal density between the cranio-caudal (CC) and medio-lateral oblique (MLO) mammographic views. Six film-readers independently re-read 59 ILC mammograms and a matched sample of 59 infiltrating ductal carcinoma (IDC) mammograms from one 3-year screening round to quantify lesion characteristics. There is fair to moderate reader agreement for parenchymal pattern, lesion shape and location (kappa = 0.41–0.60). Both ILC (33/60, 55%) and IDC (22/65, 37%) appear as a spiculate mass more often on the CC view than on the MLO view. 41% (25/60) of the ILC spiculate masses become architectural distortions or asymmetric densities on the MLO view. No more ILC lesions (4/60, 7%) are seen in dense breasts than IDC (5/65, 8%), but ILC is mainly associated with (58/60, 97%), and rarely isolated from (2/60, 3%), the main glandular density. The appearance of ILC is significantly different between the MLO and CC views (paired Wilcoxon test: z = –17.059; significance level 
0.0005). IDC appearance is not significantly different between these two views (z = –1.244; significance level 0.213). In conclusion, the CC view is optimum for distinctly visualizing ILC as a spiculate mass, as it appears as a more subtle distortion or asymmetry on the MLO view. ILC is not often isolated from the main glandular density and so optimizing visualization of this area of the breast is key to perception.
|
Introduction
|
|---|
Infiltrating lobular carcinoma (ILC) is the second most common type of breast cancer after infiltrating ductal carcinoma (IDC), accounting for between 5% and 15% of invasive breast cancers, depending on the diagnostic criteria used [1]. Its mammographic subtlety challenges the film-reader's perceptive skills, as its insidious growth pattern provokes few changes in the surrounding parenchyma, giving rise to more subtle lesion features or occult appearances in the mammographic screening scenario. As a result, ILC is more likely to present as an interval cancer between screens [2].
Previous studies regarding the mammographic appearances and location of ILC have been contradictory. When compared with other breast cancers, a spiculate mass has been reported as the more common appearance of ILC [3, 4], but others have disagreed [5, 6]. Asymmetric density and architectural distortion are the most common ILC appearances missed at screening [7, 8]. It has been observed that these subtle lesions are sometimes impossible to identify and will only be diagnosed on subsequent mammography as the cancer increases in size and begins to emerge from the background parenchyma [9].
Tumour location has been briefly described in terms of quadrant. Winchester et al [8] reported tumour location to be similar for both ILC and IDC, with 37% appearing in the upper outer quadrant; however, ILC had a slight tendency for a more central location. Tardivon et al [10], who studied all tumour types, reported that nearly half of all small tumours (<15 mm) were located in the retroglandular fat tissue or the medial half of the breast, and 18% were seen in only one view. Wang et al [11] reported that more missed lesions were located in the retroglandular area, but found no difference for medial or lateral locations.
Some authors have found the cranio-caudal (CC) view to be the best or only view for demonstrating the neo- or isodense ILC lesion [12–14]. Others have found no evidence for this microscopically, and argue that the reduced compressibility of ILC is the reason why it is better seen in the CC view [15]. Helvie [16] suggested that compression distorts both the stroma and the lesion, thus increasing its subtlety. It may simply be that ILC appears differently on each view and thus deceives the reader's perception, rather than being truly visible on one view only. This single-view appearance of ILC has been reported in the literature [5, 14]; however, as Brenner [9] pointed out, this is often a result of technical factors, e.g. the lesion being beyond the geometric limit of that view.
Young et al [17] showed that film density affects breast parenchyma visualization and adequate penetration. Earlier studies contended with lower but acceptable film densities. Tardivon et al [10] discussed this factor as one reason for false-negative mammograms and stated that fatty lobules must be identified throughout dense breast tissue.
This study set out to investigate the mammographic appearances (shape and position in relation to the breast parenchyma) of ILC by comparing the medio-lateral oblique (MLO) and CC views of all screen-detected ILC and a control group of screen-detected IDC cases from one 3-year screening round. It used six experienced film-readers to independently re-read and describe the lesion's characteristics. Descriptions of the possible shape, position and relationship of ILC to the breast parenchyma may increase awareness of these features, improve detection and therefore increase survival rates.
|
Methods and materials
|
|---|
This study was based on one 3-year screening round, from February 2002 to February 2005 using film-screen mammograms. The study population comprised 
130 000 women of screening age (50–70 years) and self-referrers over 70 years old. All cases of ILC were identified and a matched stratified sample of IDC cases was randomly selected and split across the unit's screening population of five districts. This was to ensure an equal spread over the 3-year period and to reduce possible geographical urban/rural bias.
Of the 71 cases of ILC, 12 exclusions were made. Similar exclusions affected 14 originally selected cases of IDC, and so the next randomized case was used for the study. Table 1
explains the exclusions.
The six film-readers (three radiologists and three radiographers) were all trained according to the NHS Breast Screening Programme guidelines, i.e. reading a minimum of 5000 screening mammograms per year [18] and having experience ranging from 14 months to 20 years. Ethical approval was sought from the local ethics committee, and was granted to use the patient's screening records and to anonymize images for the study. Additionally, they requested that informed consent was obtained from the film-readers, whose opinions should also be anonymous and confidential.
The mammograms were all film-based and produced using a variety of the following mammographic equipment; Mammomat 3000 (Siemens, Erlangen, Germany), Senographe DMR (GE Medical Systems, Milwaukee, WI) and Instrumentarium Alpha RT (Xograph, Tetbury, UK). A Microvision film system (Eastman Kodak, Rochester, NY) was used with a target density value of 1.75.
The 6 film-readers were asked to re-read 118 known cancer cases, and were blinded to the cancer type. A categorical observation schedule was used to record the identified lesion's features. This was divided into five sections recording breast density (using the Breast Imaging-Reporting and Data System (BI-RADS) [19]), lesion position, shape, size and any additional features seen. The categories were designed to be exclusive, but some overlap occurred because of reader interpretation. A classification guide and the team-based reading experience were expected to improve consistent performance. The results were entered into a relational database and a majority decision was taken (3+) for each section. In 16 cases, a majority decision was not possible and so these cases were re-read at a consensus meeting. Previous mammograms were not used, as the study was aimed at lesion description rather than temporal variation.
Descriptive statistics were undertaken. Kappa (
) co-efficient (PEPI version 4, 2001) was used to quantify the similarity of reader weightings for the categories used [20], and a non-parametric paired Wilcoxon test (SPSS statistical package version 14) was used to verify whether the visibility of a lesion was different on the MLO or CC view.
|
Results
|
|---|
In the screening round (February 2002 to February 2005), pure ILC represented 9% (71/781) of the total number of cancers detected. 62% (484/781) were IDC; the other 29% were classed as ductal carcinoma in situ (DCIS), mixed types and pure special types. ILC tumours tended to be larger than IDC tumours (Table 2) — the mean ILC was 23 mm in size, and the mean IDC 16 mm. 9/59 ILC cases demonstrated multifocality compared with 3/59 IDC cases. 47% (28/59) of ILC cases were treated with mastectomy compared with 25% (15/59) of IDC cases, reflecting the larger size, diffuse nature and multifocality of ILC.
The ILC cohort contained one image-detected multifocal ILC lesion and three additional image-detected multifocal IDC lesions. The IDC cohort contained three imaged-detected multifocal IDC lesions, resulting in the number of cancers analysed being 60 ILC and 65 IDC lesions.
Figure 1 shows that breast density patterns were similar for both study groups; 75% of all cases fell in the middle two BI-RADS categories using all four views. There was fair to moderate agreement between readers for categorizing density patterns (ILC, = 0.37; IDC, = 0.42). Fatty breast density pattern had the highest agreement (ILC, = 0.58; IDC, = 0.60). Fewer ILC lesions were seen in dense and heterogeneously dense breasts (ILC, n = 18; IDC, n = 26). Occult (or overlooked) lesions were spread across the density pattern types, and there were no more occult lesions in dense breast tissue than other density types (Table 3), although numbers were small. More ILC tumours were occult (ILC, n = 5; IDC, n = 1; Table 4). Only one superiorly located tumour was not on the CC view. 10 IDC tumours were not imaged, as they were situated superiorly in the breast and geometrically beyond the limits of the CC view. Therefore, 11 lesions were not present on one or other mammographic view. The two MLO-overlooked lesions were medial and found in the infra-mammary fold. Only seven cases were occult or subtle at imaging.
View larger version (15K):
[in this window]
[in a new window]
|
Figure 1. Background density pattern. IDC, infiltrating ductal carcinoma; ILC, infiltrating lobular carcinoma.
|
|
Figure 2 shows that most tumours appeared in the upper outer quadrant for both cohorts. Fair to moderate agreement was seen for all lesion positions with respect to the parenchymal density ( = 0.39–0.51). A higher proportion of IDC appeared medially (IDC, 29%; ILC, 14%) and slightly more ILC lesions appeared behind the nipple on both views (IDC, 2.5%; ILC 6%). A higher number of IDC masses were isolated from the main breast density compared with ILC masses, but numbers were low (IDC, n = 8; ILC, n = 2; Figure 3). One more ILC lesion appeared on the edge of the breast parenchyma than IDC lesions (ILC, n = 25; ILC, n = 24). However, a lesion seen on the edge of the breast parenchyma in one view may be then superimposed over the parenchyma in the other view. ILCs showed this superimposition more than IDCs (ILC, n = 13; IDC, n = 6). Stromal distortions, spiculate masses and ill-defined masses were all mainly associated within, or on the edge of, the breast density. Stromal distortion in particular will be associated with breast density where the parenchyma becomes misshapen.
View larger version (9K):
[in this window]
[in a new window]
|
Figure 2. Tumour locations in the breast by quadrant. IDC, infiltrating ductal carcinoma; ILC, infiltrating lobular carcinoma; CC, cranio-caudal; MLO, medio-lateral oblique.
|
|
View larger version (8K):
[in this window]
[in a new window]
|
Figure 3. Tumours isolated from the main breast density on each view. IDC, infiltrating ductal carcinoma; ILC, infiltrating lobular carcinoma; CC, cranio-caudal; MLO, medio-lateral oblique; IDM, ill-defined mass; SPM, speculate mass; STD, stromal/architectural distortion; ASD, asymmetric density; WDM, well-defined mass; CALC, calcification.
|
|
Figure 4 reports the tumour shapes on the CC and MLO views for both cohorts. Reader agreement for tumour shape was fair to moderate ( = 0.39–0.51). There were more ILC (n = 33, 55%) than IDC (n = 22, 37%) spiculate masses on the CC view. Of these, ILC spiculate masses became a stromal distortion, an asymmetry or occult on the MLO view (Figure 5). More ILCs presented as a stromal distortion on the MLO view (n = 15, 25%), and over half of these became spiculate masses on the CC view. IDCs did not show this variation between views; however, numbers were low. ILC presenting as asymmetry on the MLO view (n = 11, 18%) became a spectrum of shapes on the CC view and thus not many asymmetries were seen in the CC view. No ILC tumours and only one IDC tumour appeared as a well-defined mass.
View larger version (11K):
[in this window]
[in a new window]
|
Figure 4. Tumour shape on each view. IDC, infiltrating ductal carcinoma; ILC, infiltrating lobular carcinoma; CC, cranio-caudal; MLO, medio-lateral oblique; IDM, ill-defined mass; SPM, speculate mass; STD, stromal/architectural distortion; ASD, asymmetric density; WDM, well-defined mass; CALC, calcification.
|
|
View larger version (13K):
[in this window]
[in a new window]
|
Figure 5. Number of spiculate masses on the CC view appearing on the MLO view as each shape. IDC, infiltrating ductal carcinoma; ILC, infiltrating lobular carcinoma; CC, cranio-caudal; MLO, medio-lateral oblique; IDM, ill-defined mass; SPM, speculate mass; STD, stromal/architectural distortion; ASD, asymmetric density.
|
|
showed fair to moderate agreement among all categories of the schedule, although agreement was better overall for IDC than for ILC categories, indicating the variability of subtle lesion description. Again, the CC view produced stronger agreement than the MLO view for ILCs; this was reversed for IDC lesions.
The paired Wilcoxon test (z) showed that the appearance of ILCs was significantly different between the MLO and CC view (z = –17.059, significance level 0.0005). The appearance of IDCs was not significantly different (z = –1.244, significance level 0.213) between views with respect to lesion shape and position.
|
Discussion
|
|---|
Although ultrasound has proven better than mammography for detecting invasive disease, particularly small and lobular cancers in dense breasts [21], and MRI is a better predictor of size [22], mammography is the screening tool of the NHS Breast Screening Programme. The subtlety of ILC is a challenge to film-readers, as its diffuse nature can be similar to normal breast parenchyma [23]. Some authors report that subtle mammographic signs, such as stromal distortion and asymmetry, are more common for ILCs than for IDCs [24], whereas others have found no difference in mammographic appearance [8]. Veltman et al [24] report that these lesions are often detected but misinterpreted as normal or benign.
In this study, even allowing for reader variability, patterns of presentation were still apparent. The parenchymal density was determined from all four views and showed fair to moderate agreement () between readers. Although 75% of all cases fell in the middle two BI-RADS categories, it does not describe geographical spread of the density pattern. Kolb et al [25] observed that breast density may be spread out on one view but then be denser over only 25% of the orthogonal view, and thereby mask a subtle lesion. Figure 6 shows an example in which the background density covers the entire breast in the MLO view but only 50% in the lateral half of the CC view. This will influence the visibility of any lesion and make detection more difficult. Similarly, lesions that appear within the main parenchymal density on one view may then be on the edge of the parenchymal density on the orthogonal view.
View larger version (53K):
[in this window]
[in a new window]
|
Figure 6. Percentage breast difference between mammographic views.(a) Medio-lateral oblique view: glandular tissue seen spread across the whole breast. (b) Cranio-caudal view: glandular tissue is foreshortened and projected denser into the lateral portion of the breast.
|
|
Lesion visibility is also related to technical aspects of mammography. Adequately demonstrating the entire breast glandular disc is a challenge that influences the film-reader's ability to perceive abnormalities. Correct positioning of the breast aids maximum visualization of the glandular tissue. The axis of the X-ray beam needs to pass through the centre of the breast density, so that geometrically the parenchyma is evenly spread out. Any tumour hidden in the parenchyma would be more easily seen through a minimum thickness of breast density. If the breast is positioned away from the centre of the beam, then the breast densities are seen thrown together and visualization is impaired. Accurate positioning and correct exposure factors ensure that fat lobules are adequately imaged to perceive any disruption of the outline suggesting an elusive cancer [10]. The use of higher film densities in recent years (1.75 in this unit) will achieve this in denser breasts, and digital mammography will further improve lesion perception in dense breast tissue [26].
Fewer ILC than IDC lesions were seen in dense breast types, which is opposite to the findings of Fasching et al [27], who, in a mixed screening/symptomatic study, reported that ILC tumours were more likely to be found in dense breasts. However, this was a mammographic screen-detected study, which may account for the lower number of ILC lesions seen in dense breasts than in other studies that included symptomatic cohorts. In addition, the number of occult lesions present was small because the study was based on a population of screen-detected cancers. The number perceived on one view only was few, as most lesions would at least be suspected in the equivalent region on the orthogonal view. However, Figure 7 illustrates one such lesion in a fatty breast from the study.
View larger version (96K):
[in this window]
[in a new window]
|
Figure 7. A lobular carcinoma seen in a fatty breast as a speculate mass(arrow) on the CC view. (a) Medio-lateral oblique view; (b) cranio-caudal view.
|
|
More tumours appeared in the upper outer quadrant for both types, which is in accordance with the literature [8, 28, 29]; this area is in the elongated narrower part of the glandular tissue. In addition, slightly more ILCs appeared centrally, confirming the findings of Winchester et al [8]; this result indicates that ILC is more often related to the main parenchymal density. There was a higher number of IDC masses isolated from the main breast density compared with ILC masses. IDC lesions can appear denser than ILC lesions against the fatty parenchymal density, as suggested by Cherel et al [30]. These isolated ILC tumours appeared as ill-defined or spiculate masses, suggesting that an ILC tumour growing in isolation has little breast parenchyma to distort. Tardivon et al [10] reported that 45% of tumours of all types, of 15 mm or less, were retroglandular or medially placed, and therefore located away from the main parenchymal density. Although size categories are different in this study, 16% and 49% of ILC and IDC tumours, respectively, of 6–20 mm in size were isolated from the parenchymal density, demonstrating that smaller tumours are easier to perceive against a fatty background.
This study showed that ILC tumour shape varied more between views than IDC tumour shape. More ILCs appeared as spiculate masses on the CC view, becoming architectural distortion, asymmetry or occult on the MLO view. IDC spiculate masses on both views also varied more than other tumour shapes in the orthogonal view, but not to the same extent as ILC tumours. This was in agreement with Evans et al [14] (another screen-based study) and with Selinko et al [31], who found classic ILC associated with a spiculate mass and stromal distortion. Hilleran et al [12] related the different appearances between views to the insidious and flattened growth pattern of ILC lesions along planes of ligaments and connective tissue, so that the tumour is seen as either "end on" or "face on" in each view. His finding that subtle lesions were best demonstrated on the CC view is supported by our study, which has shown that ILCs are more often seen as stromal distortion and asymmetry on the MLO view (seen end on), but appear as spiculate masses on the CC view when seen face on. This indicates the lesion's flattened pattern of infiltration. Broeders et al [32] also reported that stromal distortion was a precursor to a diffuse growth pattern. The paired Wilcoxon test (z) supported the finding that the shape and location of ILCs on CC and MLO views were statistically different, whereas IDC appearances did not differ.
Harvey et al [15] hypothesized that ILC is better seen in the CC view, because greater compression can be applied and an ILC tumour will then be more easily visualized owing to its decreased compressibility. However, if the tumour is insidiously infiltrating normal tissue and blending into the background density, it will therefore be equally compressed at mammography. Helvie [16] suggests that compression distorts both the stroma and the tumour. Compression on the CC view is across the long axis of the breast disc, and therefore the distortion is likely to be a combination of greater superimposition of parenchymal structures of the tumour shape and increased density in that region. Cawson et al [6] discussed tumour planes being visualized using ultrasound, which could indicate the axis of the tumour growth plane, to assess the correct angle of view in order to best visualize the lesion's flattened shape.
This study was limited by the omission of previous mammograms, which may have aided dismissal of unchanged benign features and improved perception of subtly changing lesions on each view. Broeders et al [32] reported that ILC was visible on prior mammograms in 68% (24/35) of cases. However, the presenting lesion pattern was the straightforward aim of this study. Previous tumour subtlety could be the next stage in the analysis of a prospective cohort of ILC cases. The study was restricted to a screening population 50–70 years of age and a few older self-referrers, so younger and possibly denser parenchymal patterns were excluded. Reader variation in parenchymal density and lesion shape and size was moderate and diluted the results, although trends still emerged that could be strengthened by further studies employing computer-derived classifications, and possibly computer-aided detection. The advent of digital mammography should provide further analysis of ILC lesion structure in a three-dimensional manner, and enable the trial of new technologies, namely tomo-synthesis, to assess tumour volume and its relationship to other stromal structures.
|
Conclusions
|
|---|
ILCs show inconsistency in their shape between different mammographic views and are more distinctly seen in the CC view, where they are more likely to present as a spiculate mass compared with a vague distortion or asymmetry on the MLO view. In comparison, IDCs are seen equally well in both views, with a more consistent shape indicating its growth in all directions regardless of surrounding stroma.
Although ILC was not seen any more frequently in dense breasts when compared with IDC, which is usually isolated from the parenchymal density, the distortions of ILC were usually found within, or on the edge of, the parenchymal density.
Therefore, understanding the perceptual mammographic characteristics of ILC and the connection with the technical factors affecting perception within dense parenchyma, namely the film density and projection angle to the parenchymal density axis, may aid in earlier detection and treatment.
Received for publication August 16, 2007.
Revision received December 10, 2007.
Accepted for publication February 28, 2008.
|
References
|
|---|
- Mersin H, Yildirim E, Gülben K, Berbero
lu U. Is invasive lobular carcinoma different from invasive ductal carcinoma? Eur J Surg Oncol 2003;29:390–5.[CrossRef][Medline] - Porter PL, El-Bastawissi AY, Mandelson MT, Lin MG, Khalid N, Watney EA, et al. Breast tumour characteristics as predictors of mammographic detection: Comparison of interval- and screen-detected cancers. J Natl Cancer Inst 1999;91:2020–8.[Abstract/Free Full Text]
- Uchiyama N, Miyakawa K, Moriyama N, Kumazaki T. Radiographic features of invasive lobular carcinoma of the breast. Radiat Med 2001;19:19–25.[Medline]
- Skaane P, Skjørten F. Ultrasonographic evaluation of invasive lobular carcinoma. Acta Radiol 1999;40:369–75.[Medline]
- Cornfield EJ, Wilson ARM, Athanassiou E, Galea M, Ellis IO, Elston EW, et al. Mammographic features of invasive lobular and invasive ductal carcinoma of the breast: a comparative analysis. Br J Radiol 1995;68: 450–3.[Abstract/Free Full Text]
- Cawson JN, Law E-M, Kavangh AM. Invasive lobular carcinoma: sonographic features of cancers detected in a BreastScreen Program. Australas Radiol 2001;45:25–30.[CrossRef][Medline]
- Butler RS, Venta LA, Wiley EL, Ellis RL, Dempsey PJ, Rubin E. Sonographic evaluation of ILC. AJR Am J Roentgenol 1999;172:325–30.[Abstract/Free Full Text]
- Winchester DJ, Chang HR, Graves TA, Menck HR, Bland KI, Winchester DP. A comparative analysis of lobular and ductal carcinoma of the breast: Presentation, Treatment, and Outcomes. J Am Coll Surg 1998;186:416–22.[CrossRef][Medline]
- Brenner RJ. Asymmetric densities of the breast: strategies for imaging evaluation. Semin Roentgenol 2001;36:201–16.[CrossRef][Medline]
- Tardivon AA, Guinebretière JM, Dromain C, Vanel D. Imaging and management of nonpalpable lesions of the breast. Eur J Radiol 2002;42:2–9.[CrossRef][Medline]
- Wang J, Ting-Fang Shih T, Chien-Yao Hsu J, Yiu-Wah L. The evaluation of false negative mammography from malignant and benign breast lesions. Clin Imaging 2000;24:96–103.[CrossRef][Medline]
- Hilleran DJ, Andersson IT, Lindholm K, Linnell FS. Invasive lobular carcinoma: mammographic findings in a 10-year experience. Radiology 1991;178:149–54.[Abstract/Free Full Text]
- Newstead GM, Baute PB, Toth HK. Invasive lobular and ductal carcinoma: mammographic findings and stage at diagnosis. Radiology 1992;184:623–7.[Abstract/Free Full Text]
- Evans WP, Warren Burhenne LJ, Laurie L, O'Shaughnessy KF, Castellino RA. Invasive lobular carcinoma of the breast: mammographic characteristics and computer-aided detection. Radiology 2002;225:182–9.[Abstract/Free Full Text]
- Harvey JA, Fechner RE, Moore MM. Apparent ipsilateral decrease in breast size at mammography: a sign of infiltrating lobular carcinoma. Radiology 2000;214:883–9.[Abstract/Free Full Text]
- Helvie MA. Imaging analysis: mammography. In: Harris JR, Lippman ME, Morrow M, Osborne CK, editors. Diseases of the Breast. 3rd edition. Philadelphia, PA: Lippincott Williams & Wilkins, 2004: 131–48.
- Young K, Wallis M, Blanks R, Moss S. Influence of number of views and mammographic film density on the detection of invasive cancers: Results from the NHS Breast Screening Programme. Br J Radiol 1997;70:482–8.[Abstract]
- Liston J, Wilson R, editors. Quality Assurance Guidelines for Breast Cancer Screening Radiology. NHSBSP Publication No. 59. Sheffield, UK: NHS Cancer Screening Programmes, 2005.
- American College of Radiology. The ACR Breast Imaging Reporting and Data System (BI-RADS), 4th edition. American College of Radiology, 2003. Available from: www.acr.org [Accessed 28 December 2006].
- Cohen J. A coefficient of agreement for nominal scales. Educ Psychol Meas 1960;40:295–8.
- Benson SRC, Blue J, Judd K, Harman JE. Ultrasound is now better than mammography for the detection of invasive breast cancer. Am J Surg 2004;188:381–5.[CrossRef][Medline]
- Caramella T, Chapellier C, Ettore E, Raoust I, Champorey E, Balu-Maestro C. Value of MRI in the surgical planning of invasive lobular breast carcinoma: a prospective and a retrospective study of 57 cases: comparison with physical examination, conventional imaging, and histology. Clin Imaging 2007;31:155–61.[CrossRef][Medline]
- Molland JG, Donnellan M, Janu NC, Carmalt HL, Kennedy CW, Gillett DJ. Infiltrating lobular carcinoma – a comparison of diagnosis, management and outcome with infiltrating duct carcinoma. Breast 2004;13:389–96.[CrossRef][Medline]
- Veltman J, Boetes C, van Die L, Bult P, Blickman JG, Barentsz JO. Mammographic detection and staging of invasive lobular carcinoma. Clin Imaging 2006;30:94–8.[CrossRef][Medline]
- Kolb TM, Lichy J, Newhouse JH. Occult cancer in women with dense breasts: Detection with screening US- Diagnostic yield and tumour characteristics. Radiology 1998;207:191–9.[Abstract/Free Full Text]
- Pisano ED, Gatsonis C, Hendrick E, Yaffe M, Baum JK, Acharyya S. Diagnostic performance of digital versus film mammography for breast-cancer screening. N Engl J Med 2005;353:1773–83.[Abstract/Free Full Text]
- Fasching PA, Heusinger K, Loehberg CR, Wenkel E, Lux MP, Schrauder M, et al. Influence of mammographic density on the diagnostic accuracy of tumour size assessment and association with breast cancer tumour characteristics. European J Radiol 2006;60:398–404.[CrossRef][Medline]
- Framerino dei Malatesta M, Fiorelli M, Bandiera AF, Veneziano M, Galati GM, Porfiri L. Infiltrating lobular carcinoma of the breast (ILC): diagnostic and therapeutic aspects. Eur J Gynaecol Oncol 1995;16:36–9.[Medline]
- Crum CP, Lester SC, Cotran RS. The Female Genital System & Breast. In: Kumar V, Cotran RS, Robbins SL, editors. Basic Pathology. 7th edition. Philadelphia, PA: Saunders, 2003: 679–717.
- Cherel P, Becette V, Hagay C. Stellate images: anatomic and radiologic correlations. Eur J Radiol 2005;54:37–54.[CrossRef][Medline]
- Selinko VL, Middleton LP, Dempsey PJ. Role of sonography in diagnosing and staging invasive lobular carcinoma. J Clin Ultrasound 2004;32:323–32.[CrossRef][Medline]
- Broeders MJM, Onland-Moret NC, Rijken HJTM, Hendriks JHCL, Verbeek ALM, Holland R. Use of previous screening mammograms to identify features indicating cases that would have a possible gain in prognosis following earlier detection. Eur J Cancer 2003;39:1770–5.[CrossRef][Medline]
Top
Abstract
Introduction
