Imaging for breast cancer detection has undergone an incredible transformation in the last 30 years. The patient visiting a breast imaging department in the early 1980s would only have had mammography and ultrasound available. Today, patients have the expanded options of advanced imaging, both anatomic and molecular, which now allow for much earlier detection and clearer surgical and oncologic guidance, potentially leading to optimal outcomes.
Mammography is still the gold standard for screening given its low cost and relatively low risk. It requires tissue distortion or calcium precipitation for cancer detection, thus explaining the reason that the average size of cancer at detection by this modality is 1.5 cm.
MRI uses the altered blood flow from new capillary growth in cancerous lesions to detect them at a smaller size, but it has a well-documented incidence of high false positives.1
Molecular imaging of the breast with the rise of such tools as breast-specific gamma imaging (BSGI) and positron emission mammography (PEM) has the advantage of not being dependent on capillary growth for detection as the radiotracer can diffuse into the interstitial space and be taken up by the abnormal cells.
 |
|
The concept of PEM was first conceived in the 1990s. image/courtesy Naviscan
|
PEM is the breast application of a high-resolution PET scanner. It is also known as 3-D molecular breast imaging and represents the only tomographic molecular imaging technology for breast cancer. PEM imaging requires an injection of fluorodeoxyglucose (FDG) which is taken up by glucose avid cells, allowing for biopsy of lesions as small as 1.3 mm.2
The modality has shown promise for initial staging of breast cancer patients and for evaluating their response to therapy, recurrence, and restaging at a much higher resolution than with whole body PET.3 I've been using the technology for close to five years.
PEM advances
The concept of PEM was first conceived in the 1990s by Chris Thompson, DSc, FCCPM, and his team at McGill University in Montreal. Since that time, a number of prototype devices have undergone pre-clinical testing. PEM was introduced at RSNA in 2006 as the breast application of the first commercially available organ-specific PET scanner. The scanner has proven in-plane resolution of 1.5-2.0 mm. The mean FWHM spatial resolution is 2.4 +/- 0.3 mm in-plane at three different compression levels. Early studies showed that PEM's sensitivity and specificity for characterizing suspicious lesions were both greater than 90 percent. The system also demonstrated a positive predictive value of 92 percent for identifying index breast cancers.
Subsequent studies on secondary disease comparing PEM with MRI on the PEM system have demonstrated PEM has greater specificity than MRI; PEM and MRI have comparable high sensitivity; and PEM has a 26 percent higher positive predictive value (PPV) than MRI, thus reducing the number of unnecessary biopsies.4
The scanner has a unique instrumentation that combines coincidence imaging technology with tomosynthesis for image acquisition and a hardware configuration that is similar to mammography. It features two paddles measuring 24 cm by 16.4 cm, each of which contains a coincidence detector that measures 6 cm by 16.4 cm by 6 cm. The detectors scan across the field of view (FOV) in the direction of their 6 cm dimension to cover up to 24 cm, making the maximum FOV of the system 24 cm by 16.4 cm. They contain 2-by-2-by-13 mm lutetium yttrium orthosilicate crystals coupled to positron sensitive photomultiplier tubes.5 With the detectors oriented at 180 degrees, the scanner uses coincidence detection to produce limited angle, in-plane images.
The concept of tomosynthesis is used to acquire the images because there is only limited angle sampling due to the detectors being oriented at 180 degrees instead a full 360 degree detector ring that one sees in whole body PET. Because tomosynthesis presents an issue with anisotropic spatial resolution, two orthogonal views, craniocaudal and mediolateral oblique, are needed to achieve the required high-resolution imaging in all three dimensions. The 12 in-plane images generated by PEM have a pixel size of 1.2 mm.5 The processing of the images is performed automatically using five iterations of list mode MLEM algorithm. There is no attenuation
correction available.
Immobilization from the paddles is applied to the patient's level of tolerance and is about 50 percent the pressure of a mammogram. The paddles are mounted on an articulating arm that allows the technologist to orient the detectors vertically as well as at various angles to accommodate the required views.
For example, a mediolateral oblique view can be anywhere from 30 to 60 degrees in angle depending the patient's body habitus. One difference from mammography is the use of an adjustable chair. The patient sits for the exam as each view can take up to 10 minutes.
Interpretation by a radiologist and post-image processing such as lesion measurement and image PUV (PEM uptake value) can be accomplished on a specialized workstation. PUV is different from the SUV used in PET because the PEM does not have attenuation correction.
The determination of the PUV is accomplished in much the same way as ejection fractions in MUGA scans. One draws a region of interest around the lesion to obtain the PUVmax and a region of interest in the normal ipselateral breast parenchyma to obtain the PUVmean. The ratio of the PUVmax to the PUVmean determines the PUV, lesion-to-background ratio. An LTB/PUV of 2.0 or greater with focal localization is suggestive of malignancy warranting biopsy or BI-RADS 4 classification. LTB/PUV of less than 1.5 is considered most likely benign. An LTB/PUV of 1.5 to 2.0 is considered suspicious dependent on the architecture of the FDG uptake.
Limitations of PEM
One limitation that PEM has in its current configuration is the ability to see the chest well. The thickness of the detector paddles and the patient's body habitus can significantly impair the ability to image as far back as the chest wall.
Another limitation is that PET scans are only reimbursed in the U.S. after a pathologically proven cancer has been identified. This makes it difficult for PEM to enter in a screening realm given domestic reimbursement constraints. PEM is being adopted for screening in some
international markets.
Conclusion
PEM demonstrates excellent sensitivity and specificity for evaluating for local metastases in pre-surgical patients with biopsy-proven breast cancer, for response to neoadjuvant chemotherapy, and for restaging in patients for follow-up or possible recurrence. The one question that still remains is where in the continuum of care will PEM be best utilized.
As PEM becomes more widely used, its role should become more clearly delineated. Future improvements in the configuration of the scanner also may open new possibilities for PEM.
References
1. Bleicher RJ, Ciocca RM, Egleston BL, et al. Association of routine pretreatment magnetic resonance imaging with time to surgery, mastectomy rate, and margin status. J Am Coll Surg. 2009;209(2):180-7.
2. Kalinyak JE, Schilling K, Berg WA, et al. PET-guided breast biopsy. Breast J. 2011;17(2):143-51.
3. Comparison of positron emission mammography (PEM), whole body PET, and PET/CT in breast cancer surgical planning. Podium presentation by Wendie Berg, MD, PhD, FACR. RSNA 96th Scientific Assembly and Annual Meeting, Nov. 29, 2010.
4. Berg WA, Madsen KS, Schilling K, et al. Comparative effectiveness of positron emission mammography and MRI for presurgical planning of the ipsilateral breast in women with breast cancer. Radiology. 2011;258(1):59-72.
5. MacDonald L, Edwards J, Lewellen T, et al. Clinical imaging characteristics of the positron emission mammography camera: PEM Flex Solo II. J Nucl Med. 2009;50(10):1666-75.
Jean A. Ewing, CNMT, PET, is a PEM technologist at Texas Oncology, Plano.