Radiology
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

Published online before print July 13, 2007, 10.1148/radiol.2443061536
This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF)
Right arrow Appendix E1
Right arrow Appendix E2
Right arrow All Versions of this Article:
2443061536v1
244/3/865    most recent
Right arrow Submit a response
Right arrow Alert me when this article is cited
Right arrow Alert me when eLetters are posted
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Hsiung, P.-L.
Right arrow Articles by Connolly, J. L.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Hsiung, P.-L.
Right arrow Articles by Connolly, J. L.
(Radiology 2007;244:865-874.)
© RSNA, 2007

Technical Developments

Benign and Malignant Lesions in the Human Breast Depicted with Ultrahigh Resolution and Three-dimensional Optical Coherence Tomography1

Pei-Lin Hsiung, PhD, Darshan R. Phatak, MD, Yu Chen, PhD, Aaron D. Aguirre, SM, James G. Fujimoto, PhD, and James L. Connolly, MD

1 From the Department of Electrical Engineering and Computer Science and Research Laboratory of Electronics, Massachusetts Institute of Technology, Cambridge, Mass (P.L.H., Y.C., A.D.A., J.G.F.); and the Department of Pathology, Beth Israel Deaconess Medical Center and Harvard Medical School, 330 Brookline Ave, Boston, MA 02215 (D.R.P., J.L.C.). Received September 5, 2006; revision requested October 30; revision received December 11; final version accepted January 29, 2007. Supported by U.S. Army Medical Research Material Command Program grant DAMD 17-01-1-156, National Institutes of Health grant RO1-CA75289-10, National Science Foundation grants ECS-05-22845 and BES-05-01478, Air Force Office of Scientific Research Medical Free Electron Laser Program grant F49620-01-1-0186, the Poduska Family Foundation Fund for Innovative Research in Cancer, and through the philanthropy of Gerhard Andlinger. Address correspondence to J.L.C. (e-mail: jconnoll{at}bidmc.harvard.edu).


   ABSTRACT
TOP
 ABSTRACT
INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 ADVANCES IN KNOWLEDGE
 IMPLICATION FOR PATIENT CARE
 References
 
Institutional review board approval at the participating institutions was obtained. Informed consent was waived for this HIPAA-compliant study. The study purpose was to establish the correspondence of optical coherence tomographic (OCT) image findings with histopathologic findings to understand which features characteristic of breast lesions can be visualized with OCT. Imaging was performed in 119 specimens from 35 women aged 29–81 years with 3.5-µm axial resolution and 6-µm transverse resolution at 1.1-µm wavelength on freshly excised specimens of human breast tissue. Three-dimensional imaging was performed in 43 specimens from 23 patients. Microstructure of normal breast parenchyma, including glands, lobules, and lactiferous ducts, and stromal changes associated with infiltrating cancer were visible. Fibrocystic changes and benign fibroadenomas were identified. Imaging of ductal carcinoma in situ, infiltrating cancer, and microcalcifications correlated with corresponding histopathologic findings. OCT is potentially useful for visualization of breast lesions at a resolution greater than that of currently available clinical imaging methods.

Supplemental material:
http://radiology.rsnajnls.org/cgi/content/full/2443061536/DC1
http://radiology.rsnajnls.org/cgi/content/full/2443061536/DC2

© RSNA, 2007


   INTRODUCTION
TOP
 ABSTRACT
 INTRODUCTION
MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 ADVANCES IN KNOWLEDGE
 IMPLICATION FOR PATIENT CARE
 References
 
An estimated 212 920 new cases of invasive breast cancer and 61 980 new cases of in situ cancer were expected to occur in women in the United States in 2006 (1). Yet breast cancer mortality has declined in the last decade because of both earlier detection and improved treatment. Currently, 64% of breast cancers are diagnosed at a localized stage, which is attributed to increased use of mammography for screening, as well as awareness of breast cancer symptoms in the population (2). However, a substantial percentage of early cancers are still missed, and this finding suggests that detection methods can still be improved. Conventional needle-based biopsy techniques have a disadvantage of false-negative rates that result from sampling limitations. In addition, surgical methods for therapeutic treatment critically depend on accurate methods for assessment of microscopic resection margins. A technique with the capability of performing high-spatial-resolution, real-time, subsurface imaging could permit guidance of biopsy and intraoperative monitoring of surgical procedures and thus offer immediate information to the clinician and improve patient outcome.

Optical coherence tomography (OCT) is an emerging imaging modality that can generate micrometer-resolution, cross-sectional images of tissue microstructure in situ and in real time (3,4). OCT has a number of advantages that may make it a useful adjunct to current diagnostic modalities for breast cancer. OCT has higher spatial resolution than any currently available imaging technique used in breast cancer management. The imaging depth of OCT is limited to 2–3 mm; however, OCT can be incorporated into a wide variety of endoscopic and laparoscopic imaging devices that permit internal body imaging (5,6). OCT imaging needles also have been developed that permit imaging of a cylindric volume of tissue with up to approximately 4-mm diameter and arbitrary length (7). OCT imaging needles could be incorporated into core-biopsy devices to enable a less invasive "first look" at a tissue specimen prior to excision. Imaging can be performed over a greater tissue sampling volume than that obtained by using a typical core biopsy, and such an advantage can potentially help to reduce sampling error and trauma associated with multiple biopsy procedures. OCT catheters also may provide subsurface imaging of the epithelial microstructure that is relevant for assessment of intraductal lesions in a manner similar to that of mammary ductoscopy (8).

The capability of OCT imaging to enable identification of the normal and pathologic microstructure within the cardiovascular system, gastrointestinal tract, upper respiratory tract, genitourinary tract, and pancreatobiliary tree has been demonstrated by several groups (9–19). OCT has been investigated ex vivo in a rat model of induced breast carcinogenesis (20), and it has been used recently for visualization of human lymph node morphology in late-stage metastatic squamous cell carcinoma (21). However, to date, no results of studies have been reported about the capability of OCT to aid in the visualization of normal breast microstructure and the characteristics of pathologic findings that originate in the human breast. Thus, the purpose of our study was to establish the correspondence of OCT image findings with histopathologic findings to understand which features that are characteristic of breast lesions can be visualized by using OCT.


   MATERIALS AND METHODS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
RESULTS
 DISCUSSION
 ADVANCES IN KNOWLEDGE
 IMPLICATION FOR PATIENT CARE
 References
 
Our Health Insurance Portability and Accountability Act–compliant study was approved by the institutional review board and the Committee for the Use of Humans as Experimental Subjects at the participating institutions. Waiver of additional informed consent was granted because the consent form for the surgical procedure included consent to use tissue for research purposes. Imaging was performed on discarded portions of freshly excised surgical specimens that were not needed for diagnosis.

Specimen Selection
A total of 119 specimens from 35 women (mean age, 52 years; range, 29–81 years) were imaged within 3 hours after surgical excision. Three-dimensional (3D) imaging was performed in 43 specimens from 23 patients. Specimens were collected during a 6-month period. Fresh specimens were selected on the basis of the presence of pathologic findings that were suspected at clinical imaging or gross examination, prompt arrival in the pathology laboratory, and a large enough specimen size to allow normal and pathologic tissue to be collected without interference with routine diagnostic procedures. For specimens in which in situ pathologic findings were suspected, mammography was performed on the excised specimens to guide OCT to areas of microcalcifications.

Collected specimens consisted of both normal and pathologic tissue. Benign breast tissue that was imaged consisted of fibroadipose tissue and included mammary ducts and lobules (n = 37), tissue with fibrocystic changes (n = 8), fibroadenoma (n = 7), tissue with ductal hyperplasia (n = 6), and isolated specimens of lipoma (n = 1) and neurofibroma (n = 1). Malignant lesions that were imaged included ductal carcinoma in situ (DCIS [n = 6]) lesions, infiltrating ductal carcinoma (n = 12), infiltrating lobular carcinoma (n = 8), and infiltrating carcinoma with ductal and lobular features (n = 5). Three infiltrating ductal carcinoma specimens included regions of DCIS. Nine specimens of DCIS lesions or cancer included microcalcifications. Biopsy site changes (n = 23), fat necrosis (n = 3), and fibrosis and scarring (n = 2) were also imaged.

Imaging of Specimens and Histologic Preparation
For standard OCT systems, superluminescent diode light sources are used; these light sources are commercially available, compact, and low cost, but they have limited optical bandwidths. For this reason, most OCT studies to date have been performed with axial image resolutions of 10–15 µm at 1.3-µm wavelengths. However, OCT technology (Appendix E1 [http://radiology.rsnajnls.org/cgi/content/full/2443061536/DC1]) is rapidly evolving, and performance has improved substantially. In our study, ultrahigh-resolution OCT imaging was performed with 3.5-µm axial resolution and 6-µm transverse resolution in tissue by using a femtosecond Nd:glass laser (IC-Series diode-pumped fsNd:Glass Laser; High Q Laser Productions, Hohenems, Austria) light source at a center wavelength of 1090 nm (22). This resolution is a factor of three to four times finer than that used in previous studies performed by using standard-resolution OCT systems. This commercially available laser is compact, robust, and turnkey, thus making it suitable for future clinical studies. OCT was performed by scanning the OCT beam after the beam passed through the focusing objective (postobjective scanning) to avoid aberrations and preserve the small spot size.

Cross-sectional OCT was performed by acquiring 600–1800 axial scans (each axial scan produces a transverse pixel) over a transverse dimension of 1–2 mm. Imaging was performed at 1–2 frames per second, a speed that enabled real-time adjustment of the imaging plane. The 3D OCT data sets consisting of 200–350 individual OCT images spaced 3–5 µm apart were generated by raster scanning the beam. The 3D OCT data were visualized by using magnetic resonance imaging software for 3D visualization (Amira; Mercury Computer Systems, Chelmsford, Mass).

Because near-infrared wavelengths, which are invisible to the eye, are used with OCT, image registration was performed with a visible green guiding beam that was coincident with the OCT imaging beam. Imaging was performed without contact to the tissue. When necessary, specimens were irrigated with a medium (RPMI 1640) to prevent dehydration during imaging. After imaging, the OCT imaging plane was marked on the specimen by using two microinjections of ink to designate the imaging plane orientation. After imaging and inking, specimens were immersed in Bouin fixative for 2 seconds to fix the ink and prevent smearing. Specimens were then placed in 10% buffered formalin, routinely processed, and embedded in paraffin. Imaging, inking, and fixation of specimens were performed by one individual (P.L.H.) for all specimens. Multiple 5-µm-thick tissue sections were obtained from the registered imaging plane and stained with hematoxylin-eosin.

OCT Image and Histologic Evaluation
All OCT image evaluations were performed by using corresponding histologic findings for interpretation. Histologic findings were evaluated in consensus by two pathologists with 4 years of experience (D.R.P.) and 30 years of experience (J.L.C.) in interpreting breast pathologic findings. OCT image evaluations were performed by all authors in consensus by using matched sets of OCT images and histologic findings. Each image was evaluated to assess the capability of OCT for visualization of structures and architectural morphologic characteristics of breast tissue; included among these were the presence and appearance of glands, lobules, ducts, ductal epithelium, adipocytes, cysts, and fibrous stroma.

Structures were identified on the basis of their presence in matched histologic samples and were evaluated for relative scattering intensity and shadowing due to scattering, as well as scattering pattern and distribution. OCT images of structures in benign and malignant breast lesions were evaluated for their appearance relative to normal tissue, as well as for additional features not present in normal tissue. In addition, image rendering and visualization in the en face plane obtained from 3D OCT data were investigated for all specimens to assess the effect of focusing and signal intensity attenuation on structural visualization. Volumetric 3D OCT data sets acquired were also used to assess the utility of tracking features in three dimensions to enhance visualization of breast morphology. All original OCT image data were reduced in the axial dimension by 1.38 times to correct for the approximate index of refraction of tissue (23). Minor discrepancies between histologic findings and findings on OCT images can be attributed to tissue fixation, processing, and sectioning artifacts.


   RESULTS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
DISCUSSION
 ADVANCES IN KNOWLEDGE
 IMPLICATION FOR PATIENT CARE
 References
 
Normal Tissue
Normal fibrous stroma (Fig 1) appeared heterogeneous and highly scattering and could clearly be distinguished from surrounding adipose tissue. Fibrous stroma exhibited variations in scattering intensity, which appeared to correspond to local variations in collagen fiber density. Adipocytes appeared nonscattering to low scattering and were round or oval, with individual, well-circumscribed cell borders. Adipose tissue exhibited decreased OCT signal intensity attenuation relative to the surrounding fibrous tissue at the same depth. Consequently, regions directly beneath individual adipocytes appear to have higher signal intensity than equivalent regions at the same depth. Three-dimensional OCT imaging did not elucidate additional structural information from normal fibroadipose tissue.


Figure 1A
View larger version (144K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 1a: (a) OCT image of normal fibroadipose tissue. Fibrous stroma (F) appears heterogeneous and highly scattering, whereas adipocytes (A) appear low scattering, with individual, well-circumscribed scattering borders. (b) Histologic specimen corresponding to OCT image. (Hematoxylin-eosin stain; original magnification, x40.)

 

Figure 1B
View larger version (119K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 1b: (a) OCT image of normal fibroadipose tissue. Fibrous stroma (F) appears heterogeneous and highly scattering, whereas adipocytes (A) appear low scattering, with individual, well-circumscribed scattering borders. (b) Histologic specimen corresponding to OCT image. (Hematoxylin-eosin stain; original magnification, x40.)

 
At ultrahigh-resolution OCT (Fig 2), normal glands were visible as round structures up to 100 µm in cross section. Individual glands appeared to be lower scattering than the surrounding fibrous stroma, but they were higher scattering than and distinguishable from adipocytes. The lobular unit was visible as a cluster of glands near a draining duct. However, contrast between glandular structures and fibrous stroma was sometimes low, thus making the size, shape, and extent of individual glands difficult to determine. Sequential transverse images enabled the terminal duct lobular unit to be better identified on OCT images on the basis of 3D structure (Movie 1 [http://radiology.rsnajnls.org/cgi/content/full/2443061536/DC2]).


Figure 2A
View larger version (91K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 2a: (a, b) OCT images of a terminal duct lobular unit. On b, magnified image of the lobule in inset on a shows individual glands (G) as elliptical regions of lower scattering within the higher scattering intralobular stroma. A duct (D) is clearly visible to the right of the lobule. (c, d) Histologic specimens corresponding to OCT images. (Hematoxylin-eosin stain; original magnification, x40 [c] and x100 [d].)

 

Figure 2B
View larger version (109K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 2b: (a, b) OCT images of a terminal duct lobular unit. On b, magnified image of the lobule in inset on a shows individual glands (G) as elliptical regions of lower scattering within the higher scattering intralobular stroma. A duct (D) is clearly visible to the right of the lobule. (c, d) Histologic specimens corresponding to OCT images. (Hematoxylin-eosin stain; original magnification, x40 [c] and x100 [d].)

 

Figure 2C
View larger version (132K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 2c: (a, b) OCT images of a terminal duct lobular unit. On b, magnified image of the lobule in inset on a shows individual glands (G) as elliptical regions of lower scattering within the higher scattering intralobular stroma. A duct (D) is clearly visible to the right of the lobule. (c, d) Histologic specimens corresponding to OCT images. (Hematoxylin-eosin stain; original magnification, x40 [c] and x100 [d].)

 

Figure 2D
View larger version (143K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 2d: (a, b) OCT images of a terminal duct lobular unit. On b, magnified image of the lobule in inset on a shows individual glands (G) as elliptical regions of lower scattering within the higher scattering intralobular stroma. A duct (D) is clearly visible to the right of the lobule. (c, d) Histologic specimens corresponding to OCT images. (Hematoxylin-eosin stain; original magnification, x40 [c] and x100 [d].)

 
For ducts larger than approximately 50 µm in diameter, the ductal epithelium was visible at OCT (Fig 3) as a low-scattering layer lining the duct. Ductal epithelium was lower scattering than the surrounding fibrous stroma. Scattering intensity from ductal epithelium was similar to scattering from glands. Nonscattering ductal lumina were also visible in ducts larger than approximately 50 µm in diameter, and the basal boundary between the myoepithelial layer and stroma could be distinguished. Ducts of different diameters appeared similar in architecture and could be distinguished by their continuous and branching structure. Ducts with diameters as small as 20–25 µm could be identified on the basis of 3D architecture in sequential transverse OCT images. Smaller ducts in cross section appeared similar to glands and could only be distinguished on the basis of their 3D structure (Movie 2 [http://radiology.rsnajnls.org/cgi/content/full/2443061536/DC2]).


Figure 3A
View larger version (121K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 3a: (a, b) OCT images of a normal lactiferous duct. On b, magnified image of inset on a shows that normal ductal epithelium and myoepithelium (Ep) appear lower scattering than the surrounding fibrous stroma. The basal boundary (B) between the myoepithelium and stroma is visible. (c, d) Histologic specimens corresponding to OCT images. (Hematoxylin-eosin stain; original magnification, x40 in c and x100 in d.)

 

Figure 3B
View larger version (117K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 3b: (a, b) OCT images of a normal lactiferous duct. On b, magnified image of inset on a shows that normal ductal epithelium and myoepithelium (Ep) appear lower scattering than the surrounding fibrous stroma. The basal boundary (B) between the myoepithelium and stroma is visible. (c, d) Histologic specimens corresponding to OCT images. (Hematoxylin-eosin stain; original magnification, x40 in c and x100 in d.)

 

Figure 3C
View larger version (142K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 3c: (a, b) OCT images of a normal lactiferous duct. On b, magnified image of inset on a shows that normal ductal epithelium and myoepithelium (Ep) appear lower scattering than the surrounding fibrous stroma. The basal boundary (B) between the myoepithelium and stroma is visible. (c, d) Histologic specimens corresponding to OCT images. (Hematoxylin-eosin stain; original magnification, x40 in c and x100 in d.)

 

Figure 3D
View larger version (131K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 3d: (a, b) OCT images of a normal lactiferous duct. On b, magnified image of inset on a shows that normal ductal epithelium and myoepithelium (Ep) appear lower scattering than the surrounding fibrous stroma. The basal boundary (B) between the myoepithelium and stroma is visible. (c, d) Histologic specimens corresponding to OCT images. (Hematoxylin-eosin stain; original magnification, x40 in c and x100 in d.)

 
Benign Disease
At OCT, fibrous tissue within regions of fibrocystic change exhibited scattering characteristics (Fig 4) that were similar to characteristics of normal fibrous tissue, and individual cysts were visible as well-circumscribed structures occasionally containing scattering material. The epithelial lining of cysts was visible as a thin low-scattering band that was approximately 30–40-µm in thickness and that was similar in appearance to the normal ductal epithelium. Three-dimensional OCT did not elucidate additional structural information from breast tissue that exhibited fibrocystic changes.


Figure 4A
View larger version (114K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 4a: (a) OCT image of benign fibrocystic changes. Individual cysts (C) within fibrous stroma appear well circumscribed and irregularly filled with scattering material. Epithelium is occasionally visible as a uniform-scattering band lining cysts. (b) Histologic specimen corresponding to OCT image. (Hematoxylin-eosin stain; original magnification, x40.)

 

Figure 4B
View larger version (155K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 4b: (a) OCT image of benign fibrocystic changes. Individual cysts (C) within fibrous stroma appear well circumscribed and irregularly filled with scattering material. Epithelium is occasionally visible as a uniform-scattering band lining cysts. (b) Histologic specimen corresponding to OCT image. (Hematoxylin-eosin stain; original magnification, x40.)

 
Fibroadenoma specimens that were imaged appeared more uniformly scattering than did the heterogeneous fibrous stroma that was associated with normal fibroadipose tissue (Fig 5). Compressed ducts with epithelium were visible as fingerlike projections of low scattering. Regions of increased scattering, suggestive of boundaries between stromal nodules, were observed and could be correlated to nodule boundaries by using 3D imaging.


Figure 5A
View larger version (111K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 5a: (a) OCT image of benign fibroadenoma. Stromal nodules appear more uniformly scattering than normal fibrous tissue. Compressed ducts with epithelial component (arrowheads) were visible as fingerlike regions of low scattering. (b) Histologic specimen corresponding to OCT image. (Hematoxylin-eosin stain; original magnification, x40.)

 

Figure 5B
View larger version (161K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 5b: (a) OCT image of benign fibroadenoma. Stromal nodules appear more uniformly scattering than normal fibrous tissue. Compressed ducts with epithelial component (arrowheads) were visible as fingerlike regions of low scattering. (b) Histologic specimen corresponding to OCT image. (Hematoxylin-eosin stain; original magnification, x40.)

 
Highly scattering fingerlike projections into the duct lumen with benign ductal hyperplasia were visible (Fig 6) and were distinct from the appearance of normal ducts (Fig 3). These projections may have corresponded to fibrovascular cores that were projecting into the lumen or the boundaries between the folds of hyperplastic epithelium within the duct lumen visible at histologic evaluation. Three-dimensional imaging enhanced visualization of these projections by allowing features to be tracked in multiple image planes. Surrounding fibroadipose tissue had scattering characteristics similar to those of normal fibroadipose tissue.


Figure 6A
View larger version (89K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 6a: (a, b) OCT images of ductal hyperplasia. On b, magnified image of ductal region in inset on a shows highly scattering projections into the ductal lumen corresponding to hyperplastic epithelium or fibrovascular core (arrowheads). (c, d) Histologic specimens corresponding to OCT images. (Hematoxylin-eosin stain; original magnification, x40 in c and x100 in d.)

 

Figure 6B
View larger version (107K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 6b: (a, b) OCT images of ductal hyperplasia. On b, magnified image of ductal region in inset on a shows highly scattering projections into the ductal lumen corresponding to hyperplastic epithelium or fibrovascular core (arrowheads). (c, d) Histologic specimens corresponding to OCT images. (Hematoxylin-eosin stain; original magnification, x40 in c and x100 in d.)

 

Figure 6C
View larger version (97K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 6c: (a, b) OCT images of ductal hyperplasia. On b, magnified image of ductal region in inset on a shows highly scattering projections into the ductal lumen corresponding to hyperplastic epithelium or fibrovascular core (arrowheads). (c, d) Histologic specimens corresponding to OCT images. (Hematoxylin-eosin stain; original magnification, x40 in c and x100 in d.)

 

Figure 6D
View larger version (107K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 6d: (a, b) OCT images of ductal hyperplasia. On b, magnified image of ductal region in inset on a shows highly scattering projections into the ductal lumen corresponding to hyperplastic epithelium or fibrovascular core (arrowheads). (c, d) Histologic specimens corresponding to OCT images. (Hematoxylin-eosin stain; original magnification, x40 in c and x100 in d.)

 
Malignant Disease
In regard to OCT of DCIS lesions (Fig 7), tumor cells within the lobules appeared lower scattering than the surrounding fibrous stroma. Dilatation of the lobules caused by the tumor cells and architectural distortion of the surrounding stroma were clearly visible. Similar dilatation and distortion were visible in ducts containing DCIS lesions. In specimens with ductal hyperplasia and ductal carcinoma, distinct lumina of ducts were not visible on OCT images. Microcalcifications that were missed because they were not in the plane of the corresponding histologic micrograph were visible on the OCT image as a focal region of increased scattering with pronounced shadowing. Fibroadipose tissue surrounding the lobules with DCIS lesions had scattering characteristics similar to those of normal fibroadipose tissue. The extent of the lobules with DCIS lesions and the location of the microcalcifications could be clearly identified by examining the 3D data set (Movie 3 [http://radiology.rsnajnls.org/cgi/content/full/2443061536/DC2]).


Figure 7A
View larger version (87K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 7a: (a) OCT image of DCIS lesions in lobules. Tumor cells within lobules appear uniformly low scattering. Dilatation and architectural distortion of the lobules is visible. A microcalcification (C, circled area) within the lobules appears highly scattering with pronounced shadowing. (b) Histologic specimen corresponding to OCT image. (Hematoxylin-eosin stain; original magnification, x40.)

 

Figure 7B
View larger version (125K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 7b: (a) OCT image of DCIS lesions in lobules. Tumor cells within lobules appear uniformly low scattering. Dilatation and architectural distortion of the lobules is visible. A microcalcification (C, circled area) within the lobules appears highly scattering with pronounced shadowing. (b) Histologic specimen corresponding to OCT image. (Hematoxylin-eosin stain; original magnification, x40.)

 
OCT images of infiltrative ductal carcinoma (Fig 8) revealed highly scattering and irregular regions that appeared to correspond to tongues of cancer cells that extended into the more uniform and lower-scattering connective tissue that corresponded to desmoplastic stroma. Microcalcifications similar in appearance to those present in DCIS lesions also were visible. Three-dimensional imaging enabled tracking of tongues of cancer cells, but its primary usefulness was to aid in the identification of microcalcifications that could be missed on individual images.


Figure 8A
View larger version (111K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 8a: (a) OCT image of infiltrating ductal carcinoma. Highly scattering regions (arrowheads), which correspond to tongues of invasive cancer, are visible. (b) Histologic specimen corresponding to OCT image. (Hematoxylin-eosin stain; original magnification, x40.)

 

Figure 8B
View larger version (164K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 8b: (a) OCT image of infiltrating ductal carcinoma. Highly scattering regions (arrowheads), which correspond to tongues of invasive cancer, are visible. (b) Histologic specimen corresponding to OCT image. (Hematoxylin-eosin stain; original magnification, x40.)

 
In contrast to OCT images of infiltrating ductal carcinoma, those of infiltrative lobular carcinoma (Fig 9) appeared more homogeneously scattering. Isolated bands of slightly increased scattering appeared to correspond to intervening fibrotic stroma, and entrapped fat was visible in some specimens. On OCT images, infiltrating lobular carcinoma often was indistinguishable from regions of fibrosis, fat necrosis, and inflammatory infiltrate associated with biopsy site changes.


Figure 9A
View larger version (58K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 9a: (a) OCT image of a solid variant infiltrative lobular carcinoma. Regions with densely infiltrating tumor cells appear low scattering and homogeneous, with isolated regions of entrapped fat. (b) Histologic specimen corresponding to OCT image. (Hematoxylin-eosin stain; original magnification, x40.)

 

Figure 9B
View larger version (138K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 9b: (a) OCT image of a solid variant infiltrative lobular carcinoma. Regions with densely infiltrating tumor cells appear low scattering and homogeneous, with isolated regions of entrapped fat. (b) Histologic specimen corresponding to OCT image. (Hematoxylin-eosin stain; original magnification, x40.)

 
Three-dimensional Rendering
Figure 10 shows en face images rendered from 3D OCT data of normal fibroadipose tissue in contrast with DCIS lesions within lobules and infiltrative lobular carcinoma. Large round dilated lobules within regions of adipose tissue were visible on the image of a DCIS lesion, and they correspond to those seen in representative histologic specimens. Infiltrating lobular carcinoma with intervening fibrotic tissue more closely resembles structures visible in histologic specimens in the en face plane than structures visible in the transverse plane (Fig 9). The capability for visualization of arbitrary planes at different depths enabled 3D tissue organization to be assessed in a manner that was not possible by using individual images. Three-dimensional OCT also enabled identification of microcalcifications within a duct containing a DCIS lesion that were not initially visible on individual images (Movie 4 [http://radiology.rsnajnls.org/cgi/content/full/2443061536/DC2]).


Figure 10A
View larger version (216K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 10a: OCT images at 100 µm below the tissue surface rendered from 3D volume data sets. Three-dimensional OCT data can be viewed from a virtual en face perspective, thus yielding a view similar to that from confocal microscopy. (a) Normal breast stroma with adipocytes and fibrous tissue, (b) DCIS lesion in dilated lobules within fat, and (c) infiltrating solid variant lobular carcinoma. (d–f) Images of corresponding histologic specimens obtained in the plane perpendicular to the rendered OCT images. Each image consists of approximately 850 x 333 pixels and was rendered from 333 transverse sections. (Hematoxylin-eosin stain; original magnification, x40.)

 

Figure 10B
View larger version (234K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 10b: OCT images at 100 µm below the tissue surface rendered from 3D volume data sets. Three-dimensional OCT data can be viewed from a virtual en face perspective, thus yielding a view similar to that from confocal microscopy. (a) Normal breast stroma with adipocytes and fibrous tissue, (b) DCIS lesion in dilated lobules within fat, and (c) infiltrating solid variant lobular carcinoma. (d–f) Images of corresponding histologic specimens obtained in the plane perpendicular to the rendered OCT images. Each image consists of approximately 850 x 333 pixels and was rendered from 333 transverse sections. (Hematoxylin-eosin stain; original magnification, x40.)

 

Figure 10C
View larger version (220K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 10c: OCT images at 100 µm below the tissue surface rendered from 3D volume data sets. Three-dimensional OCT data can be viewed from a virtual en face perspective, thus yielding a view similar to that from confocal microscopy. (a) Normal breast stroma with adipocytes and fibrous tissue, (b) DCIS lesion in dilated lobules within fat, and (c) infiltrating solid variant lobular carcinoma. (d–f) Images of corresponding histologic specimens obtained in the plane perpendicular to the rendered OCT images. Each image consists of approximately 850 x 333 pixels and was rendered from 333 transverse sections. (Hematoxylin-eosin stain; original magnification, x40.)

 

Figure 10D
View larger version (213K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 10d: OCT images at 100 µm below the tissue surface rendered from 3D volume data sets. Three-dimensional OCT data can be viewed from a virtual en face perspective, thus yielding a view similar to that from confocal microscopy. (a) Normal breast stroma with adipocytes and fibrous tissue, (b) DCIS lesion in dilated lobules within fat, and (c) infiltrating solid variant lobular carcinoma. (d–f) Images of corresponding histologic specimens obtained in the plane perpendicular to the rendered OCT images. Each image consists of approximately 850 x 333 pixels and was rendered from 333 transverse sections. (Hematoxylin-eosin stain; original magnification, x40.)

 

Figure 10E
View larger version (169K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 10e: OCT images at 100 µm below the tissue surface rendered from 3D volume data sets. Three-dimensional OCT data can be viewed from a virtual en face perspective, thus yielding a view similar to that from confocal microscopy. (a) Normal breast stroma with adipocytes and fibrous tissue, (b) DCIS lesion in dilated lobules within fat, and (c) infiltrating solid variant lobular carcinoma. (d–f) Images of corresponding histologic specimens obtained in the plane perpendicular to the rendered OCT images. Each image consists of approximately 850 x 333 pixels and was rendered from 333 transverse sections. (Hematoxylin-eosin stain; original magnification, x40.)

 

Figure 10F
View larger version (218K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 10f: OCT images at 100 µm below the tissue surface rendered from 3D volume data sets. Three-dimensional OCT data can be viewed from a virtual en face perspective, thus yielding a view similar to that from confocal microscopy. (a) Normal breast stroma with adipocytes and fibrous tissue, (b) DCIS lesion in dilated lobules within fat, and (c) infiltrating solid variant lobular carcinoma. (d–f) Images of corresponding histologic specimens obtained in the plane perpendicular to the rendered OCT images. Each image consists of approximately 850 x 333 pixels and was rendered from 333 transverse sections. (Hematoxylin-eosin stain; original magnification, x40.)

 

   DISCUSSION
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
ADVANCES IN KNOWLEDGE
 IMPLICATION FOR PATIENT CARE
 References
 
In our study, OCT was shown to enable visualization of characteristic features of normal tissue, benign lesions, and malignant lesions in the human breast. Imaging was performed with 3.5-µm axial resolution and 6-µm transverse resolution, and this resolution enabled the visualization of ducts as small as 20–25 µm in diameter and the visualization of epithelium and lumina within normal ducts larger than approximately 50 µm in diameter. Normal mammary ducts and lobules could be identified within fibroadipose tissue. In contrast, lobules and ducts with DCIS lesions were associated with pronounced architectural distortion of the surrounding stroma because of dilatation with cancer cells. Lumina of ducts with ductal hyperplasia, as well as lumina of ducts containing DCIS, were not visible. Occlusion of lumina, in combination with surrounding architectural distortion within a background of otherwise normal-appearing fibroadipose tissue, could potentially, therefore, be used as a characteristic for cases of hyperplasia or of DCIS.

Presence of microcalcifications within ducts and lobules, which were also visible on OCT images, could also be used as an indirect sign of possible malignancy. Infiltrative ductal carcinoma appeared as regions of increased scattering relative to adjacent stroma. This increased scattering may result from increased nuclear material in tumor cells as compared with normal cells. This finding is consistent with findings in earlier reports that describe breast cancer cells as exhibiting increased scattering relative to the surrounding stroma (20,21). However, in our study, the low scattering associated with tumor cells in specimens with DCIS lesions and infiltrative lobular carcinoma suggests that the increased scattering in specimens with infiltrating ductal carcinoma is more likely associated with a desmoplastic reaction to infiltrating cancer than it is with malignant cellular change.

Our study also demonstrated ultrahigh-resolution 3D OCT of both benign and malignant human breast lesions. Three-dimensional OCT and rendering provided complementary information to cross-sectional OCT images and allowed tracking of features through multiple images and identification of low-contrast structures that were difficult to visualize from single images. The ability to render a tissue volume enables 3D tissue organization to be assessed. A 3D reconstructed specimen can be rotated, and arbitrary planes can be viewed from any angle, and this procedure is impossible with conventional light microscopy. OCT visualization in the en face plane also has the advantage that the entire plane of the image is within constant focus; such an advantage enables minimization of depth of focus and attenuation effects and allows higher numeric aperture focusing and higher transverse resolutions to be achieved.

Ours was a preliminary study and was limited in that it was not intended to assess the sensitivity and specificity of OCT images to help identify pathologic findings. Given the range of changes associated with normal breast tissue and benign and malignant breast lesions, imaging of larger numbers of specimens would be required to determine if the observed scattering changes can be consistently associated with neoplastic development. Presence of biopsy site changes, elastotic stroma, and scarring complicated image interpretation. Although 3D OCT data enable better visualization than do single cross-sectional images, image acquisition times used in our study were too slow for future in vivo imaging situations. Advances in OCT by using newer detection techniques enable dramatic improvements in imaging speeds (24–31). Imaging speeds of approximately 300 000 axial scans per second have been demonstrated that would enable acquisition of the 3D OCT data set in less than 1 second (25). Although OCT is limited to penetration depths of 2–3 mm, 27-gauge OCT imaging needles (7), scanning devices such as micromechanical mirrors or piezoelectric fiber scanners (32–35), and imaging devices that are based on gradient-index lenses developed for microscopy promise to enable imaging in solid organs (36). These advances promise to enable ultrahigh-resolution and 3D OCT laparoscopic or needle-based imaging in the future. Although OCT currently provides only structural information to aid identification of breast lesions, newer techniques may eventually enable imaging with enhanced image contrast or, potentially, molecular specificity (37–39).

Despite the technical challenges involved, the high incidence and mortality associated with cancers of solid organs such as the breast and prostate make research into imaging techniques for early detection and treatment worth investigation. Further investigation using larger numbers of specimens and blinded studies will be required to establish sensitivity and specificity values to determine the usefulness of OCT for identification of breast cancers and suitability for pursuing in vivo studies.


   ADVANCES IN KNOWLEDGE
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 ADVANCES IN KNOWLEDGE
IMPLICATION FOR PATIENT CARE
 References
 


   IMPLICATION FOR PATIENT CARE
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 ADVANCES IN KNOWLEDGE
 IMPLICATION FOR PATIENT CARE
References
 


   FOOTNOTES
 

Abbreviations: DCIS = ductal carcinoma in situ • OCT = optical coherence tomography • 3D = three-dimensional

J.G.F. receives royalties from intellectual property owned by Massachusetts Institute of Technology and licensed to Lightlabs Imaging, Westford, Mass, and Carl Zeiss Meditec, Dublin, Calif.

Author contributions:Guarantors of integrity of entire study, P.L.H., J.G.F.; study concepts/study design or data acquisition or data analysis/interpretation, all authors; manuscript drafting or manuscript revision for important intellectual content, all authors; manuscript final version approval, all authors; literature research, P.L.H., Y.C.; clinical studies, P.L.H., D.R.P.; experimental studies, P.L.H., Y.C., A.D.A., J.L.C.; statistical analysis, P.L.H.; and manuscript editing, P.L.H., Y.C., A.D.A., J.G.F., J.L.C.


   References
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 ADVANCES IN KNOWLEDGE
 IMPLICATION FOR PATIENT CARE
 References
 

  1. American Cancer Society. Cancer facts & figures, 2006. http://www.cancer.org/docroot/stt_0.asp. Accessed November 15, 2006.
  2. American Cancer Society. Cancer prevention and early detection facts & figures 2006. http://www.cancer.org/docroot/stt_0.asp. Accessed November 15, 2006.
  3. Huang D, Swanson EA, Lin CP, et al. Optical coherence tomography. Science 1991;254:1178–1181.[Abstract/Free Full Text]
  4. Fujimoto JG, Brezinski ME, Tearney GJ, et al. Optical biopsy and imaging using optical coherence tomography. Nat Med 1995;1:970–972.[CrossRef][Medline]
  5. Sergeev AM, Gelikonov VM, Gelikonov GV, et al. In vivo endoscopic OCT imaging of precancer and cancer states of human mucosa. Opt Expr 1997;1:432–440.
  6. Tearney GJ, Brezinski ME, Bouma BE, et al. In vivo endoscopic optical biopsy with optical coherence tomography. Science 1997;276:2037–2039.[Abstract/Free Full Text]
  7. Li XD, Chudoba C, Ko T, Pitris C, Fujimoto JG. Imaging needle for optical coherence tomography. Opt Lett 2000;25:1520–1522.[Medline]
  8. Matsunaga T, Ohta D, Misaka T, et al. Mammary ductoscopy for diagnosis and treatment of intraductal lesions of the breast. Breast Cancer 2001;8:213–221.[Medline]
  9. Brezinski ME, Tearney GJ, Bouma BE, et al. Optical coherence tomography for optical biopsy: properties and demonstration of vascular pathology. Circulation 1996;93:1206–1213.[Abstract/Free Full Text]
  10. Pitris C, Brezinski ME, Bouma BE, Tearney GJ, Southern JF, Fujimoto JG. High resolution imaging of the upper respiratory tract with optical coherence tomography: a feasibility study. Am J Respir Crit Care Med 1998;157:1640–1644.[Medline]
  11. Fujimoto JG, Boppart SA, Tearney GJ, Bouma BE, Pitris C, Brezinski ME. High resolution in vivo intra-arterial imaging with optical coherence tomography. Heart 1999;82:128–133.[Abstract/Free Full Text]
  12. Pitris C, Goodman A, Boppart SA, Libus JJ, Fujimoto JG, Brezinski ME. High-resolution imaging of gynecologic neoplasms using optical coherence tomography. Obstet Gynecol 1999;93:135–139.[Abstract/Free Full Text]
  13. Jesser CA, Boppart SA, Pitris C, et al. High resolution imaging of transitional cell carcinoma with optical coherence tomography: feasibility for evaluation of bladder pathology. Br J Radiol 1999;72:1170–1176.[Abstract]
  14. Bouma BE, Tearney GJ, Compton CC, Nishioka NS. High-resolution imaging of the human esophagus and stomach in vivo using optical coherence tomography. Gastrointest Endosc 2000;51:467–474.[CrossRef][Medline]
  15. Jackle S, Gladkova N, Feldchtein F, et al. In vivo endoscopic optical coherence tomography of the human gastrointestinal tract: toward optical biopsy. Endoscopy 2000;32:743–749.[CrossRef][Medline]
  16. Sivak MV Jr, Kobayashi K, Izatt JA, et al. High-resolution endoscopic imaging of the GI tract using optical coherence tomography. Gastrointest Endosc 2000;51:474–479.[CrossRef][Medline]
  17. Jackle S, Gladkova N, Feldchtein F, et al. In vivo endoscopic optical coherence tomography of esophagitis, Barrett's esophagus, and adenocarcinoma of the esophagus. Endoscopy 2000;32:750–755.[CrossRef][Medline]
  18. Li XD, Boppart SA, Van Dam J, et al. Optical coherence tomography: advanced technology for the endoscopic imaging of Barrett's esophagus. Endoscopy 2000;32:921–930.[CrossRef][Medline]
  19. Seitz U, Freund J, Jackle S, et al. First in vivo optical coherence tomography in the human bile duct. Endoscopy 2001;33:1018–1021.[CrossRef][Medline]
  20. Boppart SA, Luo W, Marks DL, Singletary KW. Optical coherence tomography: feasibility for basic research and image-guided surgery of breast cancer. Breast Cancer Res Treat 2004;84:85–97.[CrossRef][Medline]
  21. Luo W, Nguyen FT, Zysk AM, et al. Optical biopsy of lymph node morphology using optical coherence tomography. Technol Cancer Res Treat 2005;4:539–548.[Medline]
  22. Bourquin S, Aguirre AD, Hartl I, et al. Ultrahigh resolution real time OCT imaging using a compact femtosecond Nd:glass laser and nonlinear fiber. Opt Expr 2003;11:3290–3297.
  23. Knuttel A, Boehlau-Godau M. Spatially confined and temporally resolved refractive index and scattering evaluation in human skin performed with optical coherence tomography. J Biomed Opt 2000;5:83–92.[CrossRef][Medline]
  24. Choma MA, Sarunic MV, Yang C, Izatt JA. Sensitivity advantage of swept source and Fourier domain optical coherence tomography. Opt Expr 2003;11:2183–2189.
  25. Huber R, Wojtkowski M, Fujimoto JG, Jiang JY, Cable AE. Three-dimensional and C-mode OCT imaging with a compact, frequency swept laser source at 1300 nm. Opt Expr 2005;13:10523–10538.[CrossRef]
  26. Huber R, Wojtkowski M, Fujimoto JG. Fourier Domain Mode Locking (FDML): a new laser operating regime and applications for optical coherence tomography. Opt Expr 2006;14:3225–3237.[CrossRef]
  27. Yun SH, Tearney GJ, de Boer JF, Iftimia N, Bouma BE. High-speed optical frequency-domain imaging. Opt Expr 2003;11:2953–2963.
  28. Yun SH, Tearney GJ, Bouma BE, Park BH, de Boer JF. High-speed spectral-domain optical coherence tomography at 1.3 µm wavelength. Opt Expr 2003;11:3598–3604.
  29. Nassif NA, Cense B, Park BH, et al. In vivo high-resolution video-rate spectral-domain optical coherence tomography of the human retina and optic nerve. Opt Expr 2004;12:367–376.[CrossRef]
  30. Leitgeb RA, Drexler W, Unterhuber A, et al. Ultrahigh resolution Fourier domain optical coherence tomography. Opt Expr 2004;12:2156–2165.[CrossRef]
  31. Wojtkowski M, Srinivasan VJ, Ko TH, Fujimoto JG, Kowalczyk A, Duker JS. Ultrahigh-resolution, high-speed, Fourier domain optical coherence tomography and methods for dispersion compensation. Opt Expr 2004;12:2404–2422.[CrossRef]
  32. Pan Y, Xie H, Fedder GK. Endoscopic optical coherence tomography based on a microelectromechanical mirror. Opt Lett 2001;26:1966–1968.[Medline]
  33. Helmchen F, Fee M, Tank D, Denk W. A miniature head-mounted two-photon microscope: High-resolution brain imaging in freely moving animals. Neuron 2001;31:903–912.[CrossRef][Medline]
  34. Zara JM, Yazdanfar S, Rao KD, Izatt JA, Smith SW. Electrostatic micromachine scanning mirror for optical coherence tomography. Opt Lett 2003;28:628–630.[Medline]
  35. Liu X, Cobb MJ, Chen Y, Kimmey MB, Li XD. Rapid-scanning forward-imaging miniature endoscope for real-time optical coherence tomography. Opt Lett 2004;29:1763–1765.[CrossRef][Medline]
  36. Jung JC, Schnitzer MJ. Multiphoton endoscopy. Opt Lett 2003;28:902–904.[Medline]
  37. Rao KD, Choma MA, Yazdanfar S, Rollins AM, Izatt JA. Molecular contrast in optical coherence tomography by use of a pump-probe technique. Opt Lett 2003;28:340–342.[Medline]
  38. Yang C, Choma MA, Lamb LE, Simon JD, Izatt JA. Protein-based molecular contrast optical coherence tomography with phytochrome as the contrast agent. Opt Lett 2004;29:1396–1398.[CrossRef][Medline]
  39. Xu C, Ye J, Marks DL, Boppart SA. Near-infrared dyes as contrast-enhancing agents for spectroscopic optical coherence tomography. Opt Lett 2004;29:1647–1649.[CrossRef][Medline]




This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow