Future directions in imaging of breast diseases

Diagnostic Performance of Breast-Specific Gamma Imaging (BSGI) for Breast Cancer: Usefulness of Dual-Phase Imaging with (99m)Tc-sestamibi

PURPOSE

The aim of this study was to investigate the usefulness of breast-specific gamma imaging (BSGI) with dual-phase imaging for increasing diagnostic performance and interpreter confidence.

METHODS

We studied 76 consecutive patients (mean age: 49.3 years, range: 33-61 years) who received 925 MBq (25 mCi) (99m)Tc-sestamibi intravenously. Craniocaudal and mediolateral oblique planar images were acquired for all patients. Delayed images were obtained from all patients 1 h after tracer injection, except for patients with no definite abnormal uptake. All images were classified into four categories: group 1 (definite negative) = no definite abnormal uptake; group 2 (possible negative) = symmetrically diffuse and amorphous uptake; group 3 (possible positive) = asymmetrically mild and nodular uptake; group 4 (definite positive) = asymmetrically intense and nodular uptake. To evaluate diagnostic performance, the BSGI studies were classified as positive (group 3 or 4) or negative (group 1 or 2) for malignancy according to a visual analysis. The final diagnoses were derived from histopathological confirmation and/or imaging follow-up after at least 6 months (range: 6-14 months) by both ultrasonography and mammography.

RESULTS

The patients' ages ranged from 33 to 61 years, with an average of 49.3 years. Thirteen patients were diagnosed with malignancy, and 63 patients were diagnosed as negative for malignancy. Using early images, 43 patients were classified as group 1, 12 as group 2, 10 as group 3 and 11 as group 4. Based on early images, the sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV) and accuracy of BSGI were 77 %, 83 %, 48 %, 95 % and 82 %, respectively. Dual-phase BSGI had a sensitivity, specificity, PPV, NPV and accuracy of 69 %, 95 %, 75 %, 94 % and 91 %, respectively. The BSGI specificity was significantly higher with dual-phase imaging than with single-phase imaging (p = 0.0078), but the sensitivity did not differ significantly (p = 1.0). Based on dual-phase imaging, the sensitivity, specificity, positive predictive value, negative predictive value and accuracy of BSGI for the evaluation of US BI-RADS 4 lesions were 60 %, 86 %, 67 %, 83 % and 78 %, respectively.

CONCLUSION

Dual-phase imaging in BSGI showed good diagnostic performance and would be useful for increasing interpreter diagnostic confidence, with higher specificity, positive predictive value and accuracy for breast cancer screening as well as the differential diagnosis of breast disease compared with single-phase imaging.

Relationship between prognostic factors of breast cancer and 99mTc-sestamibi uptake in patients who underwent scintimammography: Multivariate analysis of causes of false-negative results

The complementary role of sestamibi scintimammography (SSM) in patients with breast cancer (BC) is well established. The aim of this study was to establish whether a relationship exists between sestamibi uptake, evaluated as a tumour-to-background ratio (TBR), and the main prognostic factors of BC. SSM with the measurement of TBR was performed preoperatively in 102 women (median age 57 years, range 32-81 years) who underwent curative surgery for primary BC. Final pathology showed 4 (3.9%) with pT1a, 17 (16.7%) with pT1b, 44 (43.1%) with pT1c and 37 (36.3%) with pT2 breast carcinomas. The overall sensitivity of SSM was 80.4%. An ANOVA showed significant (P<0.01) differences between the TBR of patients with G1 vs. G3 tumours, and between the TBR of those with G2 vs. G3 breast carcinomas. Moreover, there was a difference (P=0.021) between the TBR of patients (n=12, 11.8%) with CEA serum levels >10 ng/ml (2.031+/-0.420), and those with normal (n=90, 88.2%) CEA values (1.713+/-0.446), whilst no difference (P=NS) was found between patients (n=27, 26.5%) with CA 15-3 >30 U/ml (1.893+/-0.401) and those with normal (n=75, 73.5%) CA 15-3 values (1.699+/-0.462). There was a mild inverse correlation between TBR and both the oestrogen (R=0.25, P=0.011) and the progesterone receptor (R=0.23, P=0.02) rate. The logistic regression analysis showed that only size and CA 15-3 serum levels represent true independent parameters, but the function was able to predict only 11 out of 21 (52.4%) patients with false-negative SSM. TBR is independent of age and mainly correlates with the size of the tumour. There are no reliable preoperative prognostic factors that are really useful for improving SSM sensitivity in patients with small breast carcinomas.

Mixture model approach to tumor classification based on pharmacokinetic measures of tumor permeability

PURPOSE

To categorize the disease severity of mammary tumors in an animal model through the application of a novel tumor permeability mixture model within a hierarchical modeling framework.

MATERIALS AND METHODS

Thirty-six rats with mammary tumors of varying grade were imaged via dynamic contrast-enhanced (CE) MRI using albumin-(Gd-DTPA)30. Time-dependent contrast agent concentration curves for blood and tumor tissue were obtained and a mathematical model of microvascular blood-tissue exchange was developed under the hypothesis that endothelial integrity is disrupted in a manner proportional to the degree of malignancy, with benign tumors showing no disruption of the vasculature endothelium. This permeability model was incorporated into a statistical model for the benign and malignant tumor subgroups that enabled automatic subject classification. The structural and statistical models were implemented using the software Nonlinear Mixed Effects Modeling (NONMEM) to statistically separate subjects into the two subgroups.

RESULTS

Individual tumor classifications (as benign or malignant) were evaluated against the Scarff-Bloom-Richardson microscopic scoring method as applied to the tumor histology of each subject. The model-based classification resulted in 90.9% sensitivity, 92.9% specificity, and 91.7% accuracy.

CONCLUSION

Mixture model analysis provides a robust method for subject classification without user intervention and bias. Although the present results are promising, additional research is needed to further evaluate this technique for diagnostic purposes.

Occult breast cancer: scintimammography with high-resolution breast-specific gamma camera in women at high risk for breast cancer

PURPOSE

To prospectively evaluate a high-resolution breast-specific gamma camera for depicting occult breast cancer in women at high risk for breast cancer but with normal mammographic and physical examination findings.

MATERIALS AND METHODS

Institutional Review Board approval and informed consent were obtained. The study was HIPAA compliant. Ninety-four high-risk women (age range, 36-78 years; mean, 55 years) with normal mammographic (Breast Imaging Reporting and Data System [BI-RADS] 1 or 2) and physical examination findings were evaluated with scintimammography. After injection with 25-30 mCi (925-1110 MBq) of technetium 99m sestamibi, patients were imaged with a high-resolution small-field-of-view breast-specific gamma camera in craniocaudal and mediolateral oblique projections. Scintimammograms were prospectively classified according to focal radiotracer uptake as normal (score of 1), with no focal or diffuse uptake; benign (score of 2), with minimal patchy uptake; probably benign (score of 3), with scattered patchy uptake; probably abnormal (score of 4), with mild focal radiotracer uptake; and abnormal (score of 5), with marked focal radiotracer uptake. Mammographic breast density was categorized according to BI-RADS criteria. Patients with normal scintimammograms (scores of 1, 2, or 3) were followed up for 1 year with an annual mammogram, physical examination, and repeat scintimammography. Patients with abnormal scintimammograms (scores of 4 or 5) underwent ultrasonography (US), and those with focal hypoechoic lesions underwent biopsy. If no lesion was found during US, patients were followed up with scintimammography. Specific pathologic findings were compared with scintimammographic findings.

RESULTS

Of 94 women, 78 (83%) had normal scintimammograms (score of 1, 2, or 3) at initial examination and 16 (17%) had abnormal scintimammograms (score of 4 or 5). Fourteen (88%) of the 16 patients had either benign findings at biopsy or no focal abnormality at US; in two (12%) patients, invasive carcinoma was diagnosed at US-guided biopsy (9 mm each at pathologic examination).

CONCLUSION

High-resolution breast-specific scintimammography can depict small (<1-cm), mammographically occult, nonpalpable lesions in women at increased risk for breast cancer not otherwise identified at mammography or physical examination.

Breast cancer diagnosis by scintimammography: a meta-analysis and review of the literature

Scintimammography is a relatively new, non-invasive diagnostic modality in the evaluation of breast cancer. The purpose of the current study was to review the existing literature on the accuracy of scintimammography in the diagnosis of breast cancer. A search of all articles published between 1st January 1967 and 31st December 1999 was conducted. A total of 64 unique studies were selected. Each scientific paper was reviewed for scientific merit by an epidemiologist, a surgeon and a surgical resident. Assessment of scientific merit was based on a scoring scheme developed for the study. The articles included in this review reported data on a total of 5340 patients assessed for breast cancer with scintimammography. The aggregated summary estimates on these patients were sensitivity: 85.2% and specificity: 86.6%. For patients with a palpable mass the sensitivity and specificity were 87.8 and 87.5%, respectively. For patients without a palpable mass the sensitivity was 66.8% and that for specificity was 86.9%. The results of this review have shown that scintimammography may be used effectively as an adjunct to mammography and physical examination in the diagnosis of breast cancer.

Comparison between radiological and artificial neural network diagnosis in clinical screening

The imaging protocol of the UK multicentre magnetic resonance imaging study for screening in women at genetic risk of breast cancer aims to assist in detecting and diagnosing malignant breast lesions. In this paper, we evaluate a three-layer, feed-forward, backpropagation neural network as an artificial radiological classifier using receiver operating characteristic (ROC) curve analysis and compare the results with those obtained using a proposed radiological scoring system for the study which currently supplements the radiologist's clinical opinion, in comparison with histological diagnosis. Based on the 76 symptomatic cases evaluated, descriptive features scored by radiologists showed considerable overlap between benign and malignant, although some features such as irregular contours and heterogeneous enhancement were more often associated with malignant pathology. In this preliminary evaluation, ROC analysis showed that the proposed scoring scheme did not perform well, indicating further refinement is required. When all 23 features were used in the neural network, its performance was poorer than that of the scoring scheme. When only ten features were used, limited to descriptors of enhancement characteristics, the neural network performed similar to the scoring scheme. This comparison shows that the neural network approach to clinical diagnosis has considerable potential and warrants further development.

Development and validation of an algorithm for registration of serial 3D MR breast data sets

We report on the development of an algorithm to improve the registration of serial 3D MR breast images using combined global translation and rotation with locally varying parameters as geometric transformations. Several phantom and volunteer data sets were acquired and registered using mutual information as a similarity measure of the matching process. After applying a global translation by using a rigid matcher, optimum horizontal and vertical rotation angles were determined. In case of the phantom measurements, angle optimization was performed for each slice of the 3D data set of the phantom, which was deliberately shifted and rotated around different axes. In case of registration of volunteer data, optimum rotation parameters were calculated for pre-selected equidistant slices of the data set to speed up the calculation time. For slices located between and outside these support slices, the rotation angles were calculated by linear interpolation and extrapolation of the slope of the regression determined by the optimized angles of the support slices. The algorithm improves the registration of serial 3D MR data sets and represents a compromise between a rigid and an elastic 3D matching procedure.

New modalities in breast imaging: digital mammography, positron emission tomography, and sestamibi scintimammography

Digital mammography, PET, and sestamibi scintimammography are three new modalities in breast imaging. DM has advantages over film-screen mammography in image storage, retrieval, and processing and may lower the recall rate. Computer-aided detection may increase the sensitivity of mammographic screening without a substantial reduction in specificity. Whereas PET and sestambi scintimammography are not useful in breast cancer screening, PET may play a role in detecting nodal metastases and monitoring treatment response, and sestamibi scintimammography in selected cases may serve as an adjunct to conventional imaging. The cost-effectiveness of these new modalities remains to be evaluated, but all have the potential to significantly advance the diagnosis and management of women with breast cancer.

Breast tumor localization using sestamibi

PURPOSE

The sensitivity of sestamibi scanning is 85% for breast lesions that measure >or=1 cm in diameter. This detection technique complements mammography and clinical examination and can benefit patients with very dense breast tissue. An innovation in nuclear imaging uses a lead marker for localization.

MATERIALS AND METHODS

A 53-year-old woman underwent scintigraphy to clarify the indeterminate findings of a mammogram. Left breast biopsies 7 and 8 years earlier had yielded benign results. Mammography revealed a somewhat asymmetric stromal pattern, but the tissue appeared stable compared with results of previous studies. No focal abnormalities were identified. The original sestamibi breast scan revealed focally increased sestamibi uptake in the left breast. She was referred for another sestamibi scan because no radiographic or palpable abnormality correlated with the scintigraphic findings, and the lesion was believed to be nonlocalizable. Histologic examination revealed high-grade, poorly differentiated infiltrating ductal adenocarcinoma.

RESULTS

After intravenous administration of Tc-99m sestamibi, the site of the lesion was identified using a lead marker, the persistence scope, and localization needles. This facilitated surgical removal.

CONCLUSION

Using a lead marker allows placement of localization wires to guide surgical breast biopsy in patients whose lesions are visible by scintigraphy but not via mammography or palpation.

Neural network-based segmentation of dynamic MR mammographic images

The usefulness of neural networks for the classification of signal-time curves from dynamic MR mammography was recently demonstrated by our group. The multi-layer perceptron under study consists of 28 input, 4 hidden, and 3 output nodes, and was trained to classify signal-time curves into three tissue classes: "carcinoma," "benign lesion," and "parenchyma." Extending this approach, it was the aim of the present study to evaluate the performance of the developed network in the segmentation of dynamic MR mammographic images in comparison to a pixel-by-pixel two-compartment pharmacokinetic analysis. The population investigated in this pilot study comprised 15 women with suspicious lesions in the breast, which were confirmed histologically after the MR examination. The neural network classified the same areas as malignant as those which were marked as being highly suspicious by the pharmacokinetic mapping approach but with the advantage that no a priori knowledge on tissue microcirculation was needed, that computation proved to be much faster, and that it yielded a unique classification into just three tissue classes.

Classification of signal-time curves from dynamic MR mammography by neural networks

The aim of this study was to test the performance of artificial neural networks for the classification of signal-time curves obtained from breast masses by dynamic MRI. Signal-time courses from 105 parenchyma, 162 malignant, and 102 benign tissue regions were examined. The latter two groups were histopathologically verified. Four neural networks corresponding to different temporal resolutions of the signal-time curves were tested. The resolution ranges from 28 measurements with a temporal spacing of 23s to just 3 measurements taken 1.8, 3, and 10 minutes after contrast medium administration. Discrimination between malignant and benign lesions is best if 28 measurement points are used (sensitivity: 84%, specificity: 81%). The use of three measurement points results in 78% sensitivity and 76% specificity. These results correspond to values obtained by human experts who visually evaluated signal-time curves without considering additional morphologic information. All examined networks yielded poor results for the subclassification of the benign lesions into fibroadenomas and benign proliferative changes. Neural networks can computationally fast distinguish between malignant and benign lesions even when only a few post-contrast measurements are made. More precise specification of the type of the benign lesion will require incorporation of additional morphological or pharmacokinetic information.

MR dynamic enhancement of breast lesions: high temporal resolution during the first-minute versus eight-minute study

PURPOSE

To investigate the value of the early phase of MR enhancement of breast lesions.

METHOD

To study 63 breast lesions (size 5-45 mm in diameter) in 56 patients, whole-breast and lesion-targeted precontrast T1 -weighted gradient-echo 2D sequences were acquired. After intravenous injection of Gd-DTPA (0.1 mmol/Kg), four targeted scans, each every 15 seconds during the first minute (1-m), and seven whole-breast scans, each every minute up to 8 minutes (8-m), were performed. The subtraction technique was used, and percent enhancement curves were obtained. The final diagnosis was obtained by histology for 36 lesions, including 28 malignancies, and by fine-needle aspiration cytology and at least 1-year negative follow-up for the remaining 27 benign lesions.

RESULTS

Significant differences in enhancement between malignant and benign lesions were found using both techniques (p<0.0001). However the ratio between the median enhancement of malignant lesions and that of benign lesions was 6.7 (15 s), 4.8 (30 s), 4.6 (45 s), and 3.8 (60 s), descending from 4.3 to 2.5 from the second to the eighth minute. The overlap between the malignant and benign curves was 9% of the malignant range with the 1-m technique, and 50% with the 8-m technique. Three blinded observers obtained a 100% sensitivity with both techniques and a specificity of 94-97% with the 1-m technique and 83-89% with the 8-m technique.

CONCLUSION

The first minute of Gd-enhancement allows a more prominent differentiation between malignant and benign breast lesions than the following times.

Management of familial breast cancer risk

Women who are members of breast cancer families are at increased risk for breast cancer. The cloning of BRCA1 and BRCA2 has made it possible to identify mutation carriers within some of these families. Management of breast cancer risk in these families, which presents enormous challenges to patients and clinicians, is addressed. Management should begin with a full evaluation of the patient, including construction of a three-generation pedigree, ascertainment of non-genetic factors that may impact on risk, information on previous and current breast health, practice of and attitudes toward screening, and the psychosocial impact of family history on the individual. Patient priorities in risk management should be explicitly reviewed; these may include survival, cancer prevention, breast preservation, optimization of quality of life or minimization of disruption of day-to-day activities. Approaches to risk management involve screening (usually considered the mainstay), anti-estrogens, prophylactic surgery and/or lifestyle modifications. Specific gene therapy may become available in the future. Management decisions should be individualized to reflect risk levels and patient priorities and goals, within bounds that are medically and scientifically reasonable. An explicit examination of different time-frames (1, 5, 10 years) is recommended given the rapid evolution of knowledge in this area.

Automatic breast region extraction from digital mammograms for PACS and telemammography applications

High spatial resolution results in very large digital mammogram file sizes. For telemammography, and picture archiving and communication systems, the large file issue introduces technical difficulties in image transmission, storage, and display. We propose extracting the breast region from the mammogram to reduce the image file size. The challenge is on how to faithfully extract breast regions from digital mammograms generated from different types of acquisition systems that contain various imaged compositions. We report an algorithm to automatically identify the orientation of breast region and extract the breast region from mammograms. Breast regions extracted from full-field digital mammograms reduce file sizes by three to five folds.

Contrast-enhanced magnetic resonance imaging of the breast

Contrast-enhanced magnetic resonance (MR) imaging is increasingly used as a complementary diagnostic modality in breast imaging. The sensitivity of MR imaging of the breast for malignancy has consistently been reported to be excellent. The specificity has been rather variable. Study methods and imaging techniques are not standardized and there is still a great deal of uncertainty about MR imaging's place in clinical practice. Nevertheless, radiologists should be familiar with the current technique and the varying MR appearance of breast tumors to improve the accuracy of this method. This paper reviews the techniques for breast MR imaging, the pathopysiologic basis of contrast enhancement in breast tumors, and the current knowledge about detection and differentiation of breast tumors. In addition, future directions for breast MR imaging are discussed.

Elastic matching of dynamic MR mammographic images

Dynamic magnetic resonance imaging (MRI) carried out with paramagnetic contrast media has been proven to increase sensitivity and specificity in the detection of breast cancer. Due to movements of the patients and changes in the shape of the breasts during the measurement period, a coregistration (matching) of the acquired data volumes is necessary to obtain higher accuracy for the localization of lesions. In this study, an algorithm for the elastic matching of dynamic MRI volume data is presented. The approach includes automatic feature extraction along with the analysis of corresponding features between the data sets. The matching is actually slice-oriented, even though information on displacement vectors in adjacent slices is taken into account. An extension of the procedure to fully three-dimensionally (3D) matching is straight forward. Up until now, the approach has been applied to 20 dynamic MRI studies. The matching time for two image data sets with 256 x 256 x 15 voxels each was about 4 min using a PC (Pentium Pro, 200MHz).

Breast cancer screening, diagnosis, and treatment

The incidence of breast cancer in US women remains disturbingly high, and unfortunately primary care physicians still frequently encounter patients in whom the disease is suspected or, even worse, confirmed. Fortunately, however, the body of knowledge surrounding the disease has grown dramatically during the past decade, and major advances have been made in the understanding of breast cancer risk, prevention, diagnosis, and treatment. Controversies persist, particularly those concerning the screening of younger women, but consensus now exists regarding many clinical issues relevant to primary care practice. Although multidisciplinary subspecialty expertise must be made available to all women with known or suspected breast cancer, the primary care physician has an important role to play when dealing with patients with this condition. The following article focuses on what primary care practitioners need to know to expertly contribute to the diagnosis, counseling, and initial treatment of women with this disease.

Advances in breast imaging

Many exciting developments are occurring in breast imaging. Digital mammography holds the promise of telemammography and computer-aided diagnosis. Mammoscintigraphy may be helpful in identifying drug-resistant tumors before therapy. There is renewed interest in evaluating ultrasound as a potential adjunctive screening tool in women with radiographically dense breasts. Finally, contrast-enhanced magnetic resonance imaging may be used more extensively in the monitoring of tumor response to primary chemotherapy and in the preoperative workup of patients being considered for breast conservation therapy.

Breast imaging. A look at current and future technologies

Mammography remains the most practical and reliable procedure for routine breast cancer screening. It also plays an important role, along with physical examination, in evaluation of symptoms of breast disease. Among the complementary imaging techniques, breast ultrasound is the most valuable. It helps differentiate cysts from solid lesions and assists in intervention. When used in combination, clinical breast examination and mammography and, if appropriate, ultrasound, offer the best opportunity for the evaluation of the breast in symptomatic and asymptomatic women.