**The Role of Molecular Imaging Technologies in Breast Cancer Diagnosis and Management**

Anne Rosenberg, Douglas Arthur Kieper, Mark B. Williams, Nathalie Johnson and Leora Lanzkowsky *Jefferson University, Hampton University, University of Virginia, Legacy Good Samaritan, Nevada Imaging Center USA* 

## **1. Introduction**

178 Imaging of the Breast – Technical Aspects and Clinical Implication

With such a breaking point the sensitivity and specificity of electrical impedance mammography are quite high: sensitivity is 84-93%, specificity – 87-99% (according to the

Impedance mammography is in the beginning of its development. The authors sincerely hope that their modest paper will help to arouse interest of a wide range of medical

The authors would like to express their gratitude to radiologists A. Bulatov and A. Frizyuk, oncologists N. Sotskova and S. Korotkov as well as to all the staff of PKF "SIM-technika" for

Gubler E. Quantitative methods for analysis and identification of pathology. Leningrad,

Karpov A.; Korotkova M., Mumtazuddin Ahmed M., Myachin M., Tsofin Yu. Seminar on

Cherepenin V.; Karpov A., Korjenevsky A., Kornienko V., Mazaletskaya A., Mazurov D. A

Dunaeva O.; Gerasimov D., Karpov A., Machin M., Tchayev A., Tsofin Yu., Tsyplyonkov V.

Karpov A.; Korjenevsky A., Mazurov D., Mazaletskaya A. 3D Electrical Impedance Scanning

Korotkova M.; Karpov A. Procedure for assessment of the mammary gland electrical

Mumtazuddin Ahmed M. Histology functional and Clinical. HEC Islamabad, 2009

electrical impedance potential mammography. Yaroslavl, 2011, Biomedical

3D electrical impedance tomography (EIT) system for breast cancer detection.

Using Backprojection Algorithm for 3D Image Reconstruction in EIT. *World Congress on Medical Physics and Biomedical Engineering*, Munich, Germany, 2009,

of Breast Cancer. *World Congress on Medical Physics and Biomedical Engineering,* 

impedance images. *XIII international conference on electrical bio-impedance*. Graz,

data given by different authors).

their help in conducting clinical trials.

Engineering, v24/is4-6, 1996

*Chicago*, 2000, p.62

Austria, 2007.

Glants S. Medical and Biological Statistics. McGraw-Hill, 1994

*Physiological Measurement.* 2001, 22, 9-18.

Electrical Impedance Tomography. IOP, 2005

**9. Acknowledgement** 

**10. References** 

1978

researchers.

Anatomic breast imaging techniques such as mammography and ultrasound are very useful in the detection of breast cancer, but can have limited sensitivity and positive predictive value, particularly in patients with dense breasts (Kolb et al., 2002). These limitations have provided the impetus for adjunctive technologies such as nuclear medicine and PET based diagnostic imaging procedures. The nuclear medicine based technique is referred to as Breast-Specific Gamma Imaging (BSGI) or molecular breast imaging (MBI) while the positron-emission tomography (PET) based technique is referred to as Positron Emission Mammography (PEM). Both have demonstrated good results in clinical studies and are increasingly being adopted into clinical practice. Although these imaging techniques have similarities, they are different in several aspects. This chapter is designed to provide an overview of these imaging technologies and their potential roles in patient management.

#### **2. Radiotracers**

Both BSGI/MBI and PEM are physiologic imaging modalities conducted through the injection of a pharmaceutical, called a tracer, which is tagged with a radioactive isotope and the resulting molecule is called a radiotracer. Each radiotracer is designed to bind to a specific target (organ, tissue, physiologic process, cell receptor or protein) while the isotope tag emits radiation that is detected by cameras placed near the patient. The cameras provide an image of the distribution of the radiotracer tracer and thus measure a specific physiologic process in the area being imaged.

#### **2.1 Isotopes**

There are two types of radioactive isotope tags used in medical imaging: single gamma emission isotopes and positron emission isotopes. Single gamma emission isotopes release a gamma ray from the nucleus. There are a variety of single gamma isotopes used in nuclear medicine. The most common isotopes used in diagnostic imaging are referred to as lowenergy isotopes with gamma-ray energies ranging from 80 – 200 kiloelectron volts (keV). The gamma ray is a photon with sufficient energy to exit the body and be captured by specially designed detectors called gamma cameras. Positron emission isotopes emit a

The Role of Molecular Imaging Technologies in Breast Cancer Diagnosis and Management 181

perversion with most of those patients reporting a metallic taste at the time of injection. The other minor reactions are listed in table 1. More serious reactions were reported in less than 0.5% of patients and included: signs and symptoms consistent with seizure occurring shortly after administration of the agent; transient arthritis, angioedema, arrythmia, dizziness, syncope, abdominal pain, vomiting, and severe hypersensitivity characterized by dyspnea, hypotension, bradycardia, asthenia, and vomiting within two hours after a second injection of Technetium Tc99m Sestamibi. However the list of serious reactions is from the total population of patients including men and women undergoing a cardiac stress test.

> **Body System N = 673**  Body as a whole 21 (3.1%) Headache 11 (1.6%) Cardiovascular 9 (1.3%) Angina 0 (0%) ST segment changes 0 (0%) Digestive System 8 (1.2) Nausea 4 (0.6%) Special Senses 132 (19.6%) Taste Perversion 129 (19.2%) Parosmia 8 (1.2%)

PEM is a positron emission imaging technique conducted with 2-[fluorine-18] fluoro-2 deoxy-D-glucose (FDG), a modified glucose molecule with a positron-emitting isotope. Breast cancers exhibit a greater uptake of FDG than the surrounding breast tissue due to their hyperglycolytic rate. For some malignant lesions, although they possess an elevated GLUT-1 transmembrane transport function, however this does not necessarily result in increased tracer uptake (Smith, 1999). Studies have established that the uptake of FDG is primarily dependent on blood flow, the type of breast malignancy and the microstructure of the lesion (nodular vs. diffuse) (Avril et al., 2001). For example, lobular carcinoma exhibits a roughly 30% lower uptake than ductal carcinoma (Avril et al., 2001). For the PEM procedure, patients must fast 4 – 6 hours prior to the injection of FDG. It is important to note that patients with compromised glucose metabolism should have their glucose level checked prior to the administration of FDG and at least one study reports that altered glucose metabolism can affect the sensitivity of this procedure (Berg et al., 2006). FDG is administered intravenously and imaging is conducted approximately 60-90 minutes post injection in order to allow sufficient time for glucose uptake into the tissue. Patients should be requested to sit quietly in a dark, calm room to avoid manipulating the distribution of FDG. A dual-phase imaging technique may be used to improve specificity of the study

The emitted positron has a mean free path of approximately 1 mm in the breast tissue before annihilation with an electron resulting in the emission of two 511 KeV gamma rays. The random nature of the displacement between the positron and gamma ray points of origin

Table 1. Reactions to Sestamibi from 673 breast imaging patients.

(imaging at both 60 and 90 minutes post injection).

**2.1.2 Pharmaceuticals for PEM** 

positron, an antimatter particle with the same mass as an electron, but with a positive charge. This positron travels a short distance from the nucleus prior to colliding with an electron. Since the positron is a particle traveling through the tissue until this collision, the patient radiation dose associated with positron emission isotope studies is generally higher than that from single gamma emission isotope examinations. This collision results in annihilation of both particles converting the mass of the two particles into energy and producing a pair of 511 kiloelectron volt (keV) gamma rays traveling approximately 180 from each other. In positron emission tomography imaging (PET) these gamma rays exit the body and are captured by a pair of opposed gamma cameras.

The units for measuring the activity of radiotracer delivered to the patient are the Becquerel and the Curie. The Becquerel (symbol Bq) is the SI-derived unit of radioactivity. One Bq is defined as: one decay (emission) per second. The curie (symbol Ci) is a unit of radioactivity, defined as 3.7×1010 decays per second. In breast imaging, the megabecquerel (MBq) and millicurie (mCi) are the most common units used; one millicurie equals 37 Megabecquerel. These units only describe the number of decays per second for a given sample and are not to be confused with the units used to describe the radiation dose they deliver to a patient. A more detailed discussion of radiation dose is provided in Section 4.

#### **2.1.1 Pharmaceuticals for BSGI/MBI**

BSGI/MBI is a single photon imaging technique that has been conducted using a variety of imaging agents, but the most common is 99mTc-hexakis-2-methoxyisobutylisonitrile also referred to as 99mTc-Sestamibi or MIBI. MIBI is a 140 KeV gamma ray emitting isotope in a lipophilic cation molecule. It was originally cleared by the US FDA for use as a cardiac perfusion agent; breast imaging was subsequently added subsequently breast imaging was added as an indication following additional clinical studies to determine its efficacy in this application. It is injected intravenously and is retained in cells likely by electronegative cellular and mitochondrial membrane potentials (Piwnica-Worms et al., 1990). Studies show that its accumulation is roughly proportional to blood flow, desmoplastic activity and cellular proliferation and therefore it accumulates preferentially in breast cancers compared with surrounding tissues (Cutrone et al., 1998). It is a lipophilic substrate for the Pglycoprotein (Pgp), a cellular efflux pump for various compounds (Ballinger et al., 1995). Therefore, Sestamibi exhibits rapid tumor wash-in (within about 2 minutes) followed by a slow tumor washout (over the course of several hours) (Sciuto et al., 2002). Based on these factors, imaging can begin within minutes after injection and can continue for up to about 90 minutes post injection, providing ample time for all required views to be conducted before the washout cycle negatively impacts lesion-to-background tracer concentration ratio. In addition, since the level of Pgp expression correlates with tumor response to cytotoxic chemotherapy, a comparison of immediate and delayed images (4 hours post injection) may be used to quantify the radiotracer washout as a measure of Pgp expression and the probability of multi-drug resistance.

There are no known contraindications for use. Reactions to Sestamibi are generally minor according to the Cardiolite drug data sheet. In the analysis of potential reactions, 3068 patients (77% men, 22% women, and 0.7% not recorded) were documented from the cardiac clinical trials and 673 were recorded from the breast imaging trials. Of the 673 breast imaging patients, all of whom were women, the most common reported reaction was taste perversion with most of those patients reporting a metallic taste at the time of injection. The other minor reactions are listed in table 1. More serious reactions were reported in less than 0.5% of patients and included: signs and symptoms consistent with seizure occurring shortly after administration of the agent; transient arthritis, angioedema, arrythmia, dizziness, syncope, abdominal pain, vomiting, and severe hypersensitivity characterized by dyspnea, hypotension, bradycardia, asthenia, and vomiting within two hours after a second injection of Technetium Tc99m Sestamibi. However the list of serious reactions is from the total population of patients including men and women undergoing a cardiac stress test.


Table 1. Reactions to Sestamibi from 673 breast imaging patients.

## **2.1.2 Pharmaceuticals for PEM**

180 Imaging of the Breast – Technical Aspects and Clinical Implication

positron, an antimatter particle with the same mass as an electron, but with a positive charge. This positron travels a short distance from the nucleus prior to colliding with an electron. Since the positron is a particle traveling through the tissue until this collision, the patient radiation dose associated with positron emission isotope studies is generally higher than that from single gamma emission isotope examinations. This collision results in annihilation of both particles converting the mass of the two particles into energy and producing a pair of 511 kiloelectron volt (keV) gamma rays traveling approximately 180 from each other. In positron emission tomography imaging (PET) these gamma rays exit the

The units for measuring the activity of radiotracer delivered to the patient are the Becquerel and the Curie. The Becquerel (symbol Bq) is the SI-derived unit of radioactivity. One Bq is defined as: one decay (emission) per second. The curie (symbol Ci) is a unit of radioactivity, defined as 3.7×1010 decays per second. In breast imaging, the megabecquerel (MBq) and millicurie (mCi) are the most common units used; one millicurie equals 37 Megabecquerel. These units only describe the number of decays per second for a given sample and are not to be confused with the units used to describe the radiation dose they deliver to a patient. A

BSGI/MBI is a single photon imaging technique that has been conducted using a variety of imaging agents, but the most common is 99mTc-hexakis-2-methoxyisobutylisonitrile also referred to as 99mTc-Sestamibi or MIBI. MIBI is a 140 KeV gamma ray emitting isotope in a lipophilic cation molecule. It was originally cleared by the US FDA for use as a cardiac perfusion agent; breast imaging was subsequently added subsequently breast imaging was added as an indication following additional clinical studies to determine its efficacy in this application. It is injected intravenously and is retained in cells likely by electronegative cellular and mitochondrial membrane potentials (Piwnica-Worms et al., 1990). Studies show that its accumulation is roughly proportional to blood flow, desmoplastic activity and cellular proliferation and therefore it accumulates preferentially in breast cancers compared with surrounding tissues (Cutrone et al., 1998). It is a lipophilic substrate for the Pglycoprotein (Pgp), a cellular efflux pump for various compounds (Ballinger et al., 1995). Therefore, Sestamibi exhibits rapid tumor wash-in (within about 2 minutes) followed by a slow tumor washout (over the course of several hours) (Sciuto et al., 2002). Based on these factors, imaging can begin within minutes after injection and can continue for up to about 90 minutes post injection, providing ample time for all required views to be conducted before the washout cycle negatively impacts lesion-to-background tracer concentration ratio. In addition, since the level of Pgp expression correlates with tumor response to cytotoxic chemotherapy, a comparison of immediate and delayed images (4 hours post injection) may be used to quantify the radiotracer washout as a measure of Pgp expression and the

There are no known contraindications for use. Reactions to Sestamibi are generally minor according to the Cardiolite drug data sheet. In the analysis of potential reactions, 3068 patients (77% men, 22% women, and 0.7% not recorded) were documented from the cardiac clinical trials and 673 were recorded from the breast imaging trials. Of the 673 breast imaging patients, all of whom were women, the most common reported reaction was taste

body and are captured by a pair of opposed gamma cameras.

more detailed discussion of radiation dose is provided in Section 4.

**2.1.1 Pharmaceuticals for BSGI/MBI** 

probability of multi-drug resistance.

PEM is a positron emission imaging technique conducted with 2-[fluorine-18] fluoro-2 deoxy-D-glucose (FDG), a modified glucose molecule with a positron-emitting isotope. Breast cancers exhibit a greater uptake of FDG than the surrounding breast tissue due to their hyperglycolytic rate. For some malignant lesions, although they possess an elevated GLUT-1 transmembrane transport function, however this does not necessarily result in increased tracer uptake (Smith, 1999). Studies have established that the uptake of FDG is primarily dependent on blood flow, the type of breast malignancy and the microstructure of the lesion (nodular vs. diffuse) (Avril et al., 2001). For example, lobular carcinoma exhibits a roughly 30% lower uptake than ductal carcinoma (Avril et al., 2001). For the PEM procedure, patients must fast 4 – 6 hours prior to the injection of FDG. It is important to note that patients with compromised glucose metabolism should have their glucose level checked prior to the administration of FDG and at least one study reports that altered glucose metabolism can affect the sensitivity of this procedure (Berg et al., 2006). FDG is administered intravenously and imaging is conducted approximately 60-90 minutes post injection in order to allow sufficient time for glucose uptake into the tissue. Patients should be requested to sit quietly in a dark, calm room to avoid manipulating the distribution of FDG. A dual-phase imaging technique may be used to improve specificity of the study (imaging at both 60 and 90 minutes post injection).

The emitted positron has a mean free path of approximately 1 mm in the breast tissue before annihilation with an electron resulting in the emission of two 511 KeV gamma rays. The random nature of the displacement between the positron and gamma ray points of origin

The Role of Molecular Imaging Technologies in Breast Cancer Diagnosis and Management 183

these detector systems is to reconstruct the gamma ray emissions into an image that allows the physician to visualize the distribution of the radiotracer in the body. There are two modes of image reconstruction, plannar and tomographic. The planar method results in a single 2-dimensional image per acquisition, similar to the mammogram. Tomographic reconstruction provides a 3-dimensional reconstruction of the breast similar to MRI. The single gamma detector systems used for BSGI/MBI can be planar or tomographic and can be constructed of a single or multiple detectors. The positron imaging systems used for PEM by their design provide tomographic imaging only and since the detection of the pair of gamma rays is required for image reconstruction, positron systems must consist of either a

The detection of an abnormality in BSGI and PEM imaging is based on the ability of the imaging system to depict the variations of uptake in the tissue. Unlike anatomical imaging where high spatial resolution is needed to visualize the detailed morphology used to provide differential diagnosis, the molecular imaging system must provide sufficient image contrast in order to visualize the variations in radiotracer uptake; the contrast between radiotracer concentration in the tumor and the uptake of the surrounding breast tissue. While this is partially a function of resolution, there are several other factors impacting imaging. This contrast based imaging requires a careful balance between spatial resolution, image noise and photon sensitivity. Generally, as spatial resolution increases, image noise increases and photon sensitivity decreases proportionally to some degree thus if spatial resolution is increased to a level where the detector has poor photon sensitivity and the resulting image noise is too high, the ability of the system to visualize the contrasting tissue uptake is diminished. A detailed discussion of the balance between these factors is beyond the scope of this text, but it is important to realize that in molecular imaging is a contrast based imaging and spatial resolution is not the only parameter affecting the visualization of lesions. For example, it is possible to detect a 1 mm cancer using a system with a 4 mm spatial resolution if the uptake of that lesion is sufficiently enough higher than the background to overcome the partial volume effect. Conversely, a 40 mm cancer could be missed by the same imaging system if the lesion uptake is not sufficiently higher than the

As an illustration, nearly all commercially available large field-of-view gamma cameras, typical to the nuclear medicine department, have a variable matrix setting, including 512 x 512, 256 x 256 and 128 x 128. Although the 512 x 512 setting produces the highest spatial resolution, nearly all nuclear medicine studies are conducted on the 256 x 256 or 128 x 128 settings because the resulting image noise at the 512 x 512 setting diminishes image quality

System photon sensitivity is another important parameter in BSGI/MBI and PEM imaging. As photon sensitivity increases, the amount of radiotracer, the length of time the image is acquired, or some combination of both parameters can be decreased. For Example: a given detector system is providing good clinical images using a dose of 300 MBq and an acquisition time of 10 minutes. If the photon sensitivity of this system can be increased 50% the clinician would have three possible options. First, they could reduce the patient's radiation exposure by reducing the amount of radiotracer delivered to the patient by 50%, to

pair of opposed detectors or a ring detector design.

**3.1 Gamma-ray imaging basics** 

surrounding tissue.

for the majority of studies.

has some impact on the lower limit of spatial resolution in studies using positron-emitting isotopes (Turkington, 2001).

#### **2.2 Comparison of BSGI/MBI and PEM Radiotracers**

Both Sestamibi and FDG demonstrate increased accumulation in breast cancers, although the mechanism for this accumulation is better understood for FDG. In addition, although the breast tissue typically has a homogeneous uptake of both tracers, they can accumulate in normal glandular tissue resulting in a diffuse heterogeneous uptake pattern, especially in pre-menopausal women who are in the luteal phase of their menstrual cycle (Lin et al., 2007). This is not surprising as it reflects the heterogeneous nature of the breast tissue and the impact of blood hormone levels on the breast parenchyma. This heterogeneity does not generally impact cancer detection, but may complicate interpretation. The intensity of this pattern can be reduced for both tracers by imaging outside of the luteal phase and several studies report that day 2 – 14 of the menstrual cycle is optimal. Neither tracer is linked to nephrogenic systemic fibrosis, a sometimes fatal condition that is associated with gadolinium contrast agents used in breast MRI. MIBI has some minor pharmacologic considerations and rare reactions occurring in less than 0.5%of patients. FDG imaging requires fasting for a minimum of 4 hours prior to injection and as mentioned in the previous section, it is beneficial to check the blood glucose level prior to FDG administration as the tumor uptake of FDG is reduced in hyperglycemic states (Schelbert et al., 1999) and this results in some potential for complications and reduced sensitivity for the procedure in diabetic patients. In comparison, MIBI does not require fasting and imaging can be conducted within minutes of the injection.

Some of the physical and clinical differences between FDG and MIBI imaging are summarized in Tables 2 and 3 respectively.


Table 2. gamma-ray emission information for radiotracers.


Table 3. Radiotracer administration and imaging considerations.

#### **3. Imaging systems**

Both the single gamma and positron isotopes described in section 2.1 ultimately emit gamma rays (the positron isotopes through the annihilation and the conversion of the positron) that exit the patient's body and can be detected with external detectors. The goal of these detector systems is to reconstruct the gamma ray emissions into an image that allows the physician to visualize the distribution of the radiotracer in the body. There are two modes of image reconstruction, plannar and tomographic. The planar method results in a single 2-dimensional image per acquisition, similar to the mammogram. Tomographic reconstruction provides a 3-dimensional reconstruction of the breast similar to MRI. The single gamma detector systems used for BSGI/MBI can be planar or tomographic and can be constructed of a single or multiple detectors. The positron imaging systems used for PEM by their design provide tomographic imaging only and since the detection of the pair of gamma rays is required for image reconstruction, positron systems must consist of either a pair of opposed detectors or a ring detector design.

#### **3.1 Gamma-ray imaging basics**

182 Imaging of the Breast – Technical Aspects and Clinical Implication

has some impact on the lower limit of spatial resolution in studies using positron-emitting

Both Sestamibi and FDG demonstrate increased accumulation in breast cancers, although the mechanism for this accumulation is better understood for FDG. In addition, although the breast tissue typically has a homogeneous uptake of both tracers, they can accumulate in normal glandular tissue resulting in a diffuse heterogeneous uptake pattern, especially in pre-menopausal women who are in the luteal phase of their menstrual cycle (Lin et al., 2007). This is not surprising as it reflects the heterogeneous nature of the breast tissue and the impact of blood hormone levels on the breast parenchyma. This heterogeneity does not generally impact cancer detection, but may complicate interpretation. The intensity of this pattern can be reduced for both tracers by imaging outside of the luteal phase and several studies report that day 2 – 14 of the menstrual cycle is optimal. Neither tracer is linked to nephrogenic systemic fibrosis, a sometimes fatal condition that is associated with gadolinium contrast agents used in breast MRI. MIBI has some minor pharmacologic considerations and rare reactions occurring in less than 0.5%of patients. FDG imaging requires fasting for a minimum of 4 hours prior to injection and as mentioned in the previous section, it is beneficial to check the blood glucose level prior to FDG administration as the tumor uptake of FDG is reduced in hyperglycemic states (Schelbert et al., 1999) and this results in some potential for complications and reduced sensitivity for the procedure in diabetic patients. In comparison, MIBI does not require fasting and imaging can be

Some of the physical and clinical differences between FDG and MIBI imaging are

**Energy of Emission photon** 

**FDG** 511 KeV 110 **MIBI** 140 KeV 360

**Imaging time** 

Both the single gamma and positron isotopes described in section 2.1 ultimately emit gamma rays (the positron isotopes through the annihilation and the conversion of the positron) that exit the patient's body and can be detected with external detectors. The goal of

**Half Life (minutes)** 

minutes Blood glucose

**Recommended Pre-procedural testing** 

isotopes (Turkington, 2001).

**2.2 Comparison of BSGI/MBI and PEM Radiotracers** 

conducted within minutes of the injection.

summarized in Tables 2 and 3 respectively.

**3. Imaging systems** 

Table 2. gamma-ray emission information for radiotracers.

**FDG** 4 - 6 hours 60 – 90

Table 3. Radiotracer administration and imaging considerations.

**Pre-study Fasting Injection to** 

**MIBI** None 5 minutes None

The detection of an abnormality in BSGI and PEM imaging is based on the ability of the imaging system to depict the variations of uptake in the tissue. Unlike anatomical imaging where high spatial resolution is needed to visualize the detailed morphology used to provide differential diagnosis, the molecular imaging system must provide sufficient image contrast in order to visualize the variations in radiotracer uptake; the contrast between radiotracer concentration in the tumor and the uptake of the surrounding breast tissue. While this is partially a function of resolution, there are several other factors impacting imaging. This contrast based imaging requires a careful balance between spatial resolution, image noise and photon sensitivity. Generally, as spatial resolution increases, image noise increases and photon sensitivity decreases proportionally to some degree thus if spatial resolution is increased to a level where the detector has poor photon sensitivity and the resulting image noise is too high, the ability of the system to visualize the contrasting tissue uptake is diminished. A detailed discussion of the balance between these factors is beyond the scope of this text, but it is important to realize that in molecular imaging is a contrast based imaging and spatial resolution is not the only parameter affecting the visualization of lesions. For example, it is possible to detect a 1 mm cancer using a system with a 4 mm spatial resolution if the uptake of that lesion is sufficiently enough higher than the background to overcome the partial volume effect. Conversely, a 40 mm cancer could be missed by the same imaging system if the lesion uptake is not sufficiently higher than the surrounding tissue.

As an illustration, nearly all commercially available large field-of-view gamma cameras, typical to the nuclear medicine department, have a variable matrix setting, including 512 x 512, 256 x 256 and 128 x 128. Although the 512 x 512 setting produces the highest spatial resolution, nearly all nuclear medicine studies are conducted on the 256 x 256 or 128 x 128 settings because the resulting image noise at the 512 x 512 setting diminishes image quality for the majority of studies.

System photon sensitivity is another important parameter in BSGI/MBI and PEM imaging. As photon sensitivity increases, the amount of radiotracer, the length of time the image is acquired, or some combination of both parameters can be decreased. For Example: a given detector system is providing good clinical images using a dose of 300 MBq and an acquisition time of 10 minutes. If the photon sensitivity of this system can be increased 50% the clinician would have three possible options. First, they could reduce the patient's radiation exposure by reducing the amount of radiotracer delivered to the patient by 50%, to

The Role of Molecular Imaging Technologies in Breast Cancer Diagnosis and Management 185

those in the superior portion of the breast. This resolution loss is, at least in part, the driving logic behind the dual-head opposed detector design. In theory, if the breast is imaged using a dual-head system in the cranio-caudal position, the upper detector would maximize visualization of the superior portion of the breast while the lower detector optimizes

Although there is a theoretical benefit to the dual-head design, it is interesting to note that clinical data from the dual-head and single-head systems shows similar performance in terms of lesion sensitivity (see section 5 below). This is likely due to the two-view imaging protocol adopted from mammography that is standard in BSGI/MBI protocols. Just as in mammography, the optimal coverage of breast tissue is obtained by acquiring an MLO and CC image of each breast and additional images are obtained as needed. Since all patients have a minimum of two views obtained, the likelihood of a lesion being deep to both projections is quite small. In addition, similar to mammography, when a lesion is seen in only one image, additional images are obtained in order to determine the location of the lesion in the breast tissue. As long as this two-view protocol remains the standard, it is unlikely that the dual-head system will result in significant improvements in the sensitivity of the detection of breast malignancies. However, provided the two detector images are fused properly, it may be possible to reduce either the injected dose or the acquisition time

Single photon emission computed tomography (SPECT) is a recent development in BSGI/MBI. Currently these devices are only available in the research setting (Williams et al., 2010). Additional research is needed to determine if dual-head image combination techniques or the implementation of SPECT imaging will provide a clear benefit to BSGI/MBI in terms of breast cancer detection. Such studies are underway, but the data is

 Fig. 2. Left and Center - a single-head BSGI/MBI system with a compression paddle used for positioning. The left image illustrates the cranio-caudal (CC) position and the center image illustrates the medial lateral oblique position (MLO). The right image illustrates the MLO position with the dual-head system where the compression paddle is replaced by a

to facilitate low-dose imaging or higher throughput on the imaging system.

visualization of inferior lesions.

not yet available for analysis.

second detector.

150 MBq and maintain the same imaging time, 10 minutes. Second, they could reduce the imaging time by 50% to 5 minutes using a 300 MBq dose. Or third, they could reduce both the time and the dose by roughly 25% resulting in a dose of 225 MBq and an imaging time of 7.5 minutes. It is important to remember that in molecular imaging techniques such as BSGI/MBI and PEM, the imaging time and the dose delivered can be manipulated, but reducing both to any large degree is not possible unless significant improvements in photon sensitivity are obtained.

## **3.1.1 BSGI/MBI Imaging**

BSGI/MBI imaging is conducted with a single-head or dual-head detector system (see Figure 1). Only one detector equipped with a collimator is required for image reconstruction. Generally, gentle breast compression (normally less than 12 lbs or 53 newtons) is used to provide breast immobilization. This compression is noticeably lower than that used in mammography for two reasons. First, the typical imaging time for a single BSGI/MBI image is significantly longer than that needed for a mammographic projection, 5 – 10 minutes, thus lower pressures are better tolerated by patients and second, the 140keV gamma ray emitted in BSGI/MBI has sufficiently higher tissue penetration than the 8 – 35 keV x-ray used in mammography therefore these images benefit less from higher compression. As shown in figure 1, in the single-head design a paddle is used to provide compression and in the dual-head system, the breast is compressed between the detectors. The compression paddle used in the single head system can be exchanged for a fenestrated paddle to allow biopsy. Biopsy is currently not available on the dual-head design.

Fig. 1. A single and dual head imaging system for BSGI/MBI.

#### **3.1.2 BSGI/MBI detectors**

As mentioned in the previous section, there is generally an inverse relationship between photon sensitivity and spatial resolution however both are important to imaging. These parameters are determined by several aspects of the detector design, especially that of the collimator. The most commercially available systems have an extrinsic spatial resolution of between 1.9 and 3.3 mm at the surface of the detector however it is important to note that the spatial resolution of planar, single gamma imaging systems decreases with increasing source-to-collimator distance, thus the spatial resolution of a lesion near the detector is better than that of one deep in the breast tissue, relative to the detector face. For example, if the breast is being imaged in the cranio-caudal position (detector inferior to the breast tissue), lesions in the inferior portion of the breast tissue will be somewhat more visible than

150 MBq and maintain the same imaging time, 10 minutes. Second, they could reduce the imaging time by 50% to 5 minutes using a 300 MBq dose. Or third, they could reduce both the time and the dose by roughly 25% resulting in a dose of 225 MBq and an imaging time of 7.5 minutes. It is important to remember that in molecular imaging techniques such as BSGI/MBI and PEM, the imaging time and the dose delivered can be manipulated, but reducing both to any large degree is not possible unless significant improvements in photon

BSGI/MBI imaging is conducted with a single-head or dual-head detector system (see Figure 1). Only one detector equipped with a collimator is required for image reconstruction. Generally, gentle breast compression (normally less than 12 lbs or 53 newtons) is used to provide breast immobilization. This compression is noticeably lower than that used in mammography for two reasons. First, the typical imaging time for a single BSGI/MBI image is significantly longer than that needed for a mammographic projection, 5 – 10 minutes, thus lower pressures are better tolerated by patients and second, the 140keV gamma ray emitted in BSGI/MBI has sufficiently higher tissue penetration than the 8 – 35 keV x-ray used in mammography therefore these images benefit less from higher compression. As shown in figure 1, in the single-head design a paddle is used to provide compression and in the dual-head system, the breast is compressed between the detectors. The compression paddle used in the single head system can be exchanged for a fenestrated

paddle to allow biopsy. Biopsy is currently not available on the dual-head design.

As mentioned in the previous section, there is generally an inverse relationship between photon sensitivity and spatial resolution however both are important to imaging. These parameters are determined by several aspects of the detector design, especially that of the collimator. The most commercially available systems have an extrinsic spatial resolution of between 1.9 and 3.3 mm at the surface of the detector however it is important to note that the spatial resolution of planar, single gamma imaging systems decreases with increasing source-to-collimator distance, thus the spatial resolution of a lesion near the detector is better than that of one deep in the breast tissue, relative to the detector face. For example, if the breast is being imaged in the cranio-caudal position (detector inferior to the breast tissue), lesions in the inferior portion of the breast tissue will be somewhat more visible than

Fig. 1. A single and dual head imaging system for BSGI/MBI.

sensitivity are obtained.

**3.1.1 BSGI/MBI Imaging** 

**3.1.2 BSGI/MBI detectors** 

those in the superior portion of the breast. This resolution loss is, at least in part, the driving logic behind the dual-head opposed detector design. In theory, if the breast is imaged using a dual-head system in the cranio-caudal position, the upper detector would maximize visualization of the superior portion of the breast while the lower detector optimizes visualization of inferior lesions.

Although there is a theoretical benefit to the dual-head design, it is interesting to note that clinical data from the dual-head and single-head systems shows similar performance in terms of lesion sensitivity (see section 5 below). This is likely due to the two-view imaging protocol adopted from mammography that is standard in BSGI/MBI protocols. Just as in mammography, the optimal coverage of breast tissue is obtained by acquiring an MLO and CC image of each breast and additional images are obtained as needed. Since all patients have a minimum of two views obtained, the likelihood of a lesion being deep to both projections is quite small. In addition, similar to mammography, when a lesion is seen in only one image, additional images are obtained in order to determine the location of the lesion in the breast tissue. As long as this two-view protocol remains the standard, it is unlikely that the dual-head system will result in significant improvements in the sensitivity of the detection of breast malignancies. However, provided the two detector images are fused properly, it may be possible to reduce either the injected dose or the acquisition time to facilitate low-dose imaging or higher throughput on the imaging system.

Single photon emission computed tomography (SPECT) is a recent development in BSGI/MBI. Currently these devices are only available in the research setting (Williams et al., 2010). Additional research is needed to determine if dual-head image combination techniques or the implementation of SPECT imaging will provide a clear benefit to BSGI/MBI in terms of breast cancer detection. Such studies are underway, but the data is not yet available for analysis.

Fig. 2. Left and Center - a single-head BSGI/MBI system with a compression paddle used for positioning. The left image illustrates the cranio-caudal (CC) position and the center image illustrates the medial lateral oblique position (MLO). The right image illustrates the MLO position with the dual-head system where the compression paddle is replaced by a second detector.

The Role of Molecular Imaging Technologies in Breast Cancer Diagnosis and Management 187

One limitation to PEM image reconstruction is that the detector photon sensitivity is not linear across the field-of-view with lower sensitivity along the detector edges. This causes a higher level of noise to be present in the breast images, along the chest wall. Figure 5 provides a schematic representation of factors affecting the photon sensitivity in a PEM detector. The maximum angle of reconstruction (MAR) is a setting used in PEM software and it is defined as the maximum angle away from the detector normals (90 from the detector face) for which coincident gamma ray detections are included in tomographic image reconstruction. Larger MAR values yield greater overall photon sensitivity but with potential loss in spatial resolution due to the depth-of-interaction (DOI) blur. The blue lines in Figure 1 show the angular range over which gamma rays emitted from two points in the breast are accepted. Figure 1A shows a point near the nipple, and Figure 1B shows one near the chest. For events that occur in the center of the field-of-view (FOV), all of the events within the MAR are captured by the detector system. However a significant fraction of events occurring near the FOV edges go uncounted for because one of the paired gamma rays traveling outside the edges of the detector is not detected. As a recent study found, this loss of photon sensitivity along the edges limits the ability of the PEM system to detect

Fig. 5. A and B: a schematic example of the maximum angle of reconstruction near the center

Detector 1 Detector 1

Detector 2 Detector 2

Fig. 6. a typical PEM study with multiple slice reconstruction for each projection.

of the detector field-of-view and then near the chest wall respectively.

lesions located near the chest wall (Rosen et al, 2005).

Fig. 3. A typical 4-view BSGI study.

#### **3.1.3 Detectors for PEM**

PEM imaging is conducted with either a dual head or ring style gamma-ray detector. Both systems are designed to detect the coincident gamma rays which, are traveling approximately 180° from each other after the annihilation reaction (see figure 4). Unlike BSGI imaging devices, PEM devices do not use a collimator to help determine the location of each event. In PEM imaging, since there are two gamma-rays traveling 180° apart, the event location is calculated as a line of response between the location that each gamma-ray strikes the pair of opposed detectors. One advantage to PEM is that it does not have the same loss of resolution with distance that BSGI/MBI systems experience. As mentioned in the previous section, the mean free path of the 18F positron within the breast tissue is approximately 1 mm and commercially available PEM systems report an in-plane spatial resolution of about 2 mm. One limitation of the dual-head PEM detector design is that, due to the limited angle of acquisition, it has limited resolution in the Z-axis (depth). Ring detectors do not suffer from this limitation as they provide a 360° acquisition for reconstruction however there is currently no biopsy capability on the ring detector systems. A needle biopsy localization device was recently introduced for the opposed dual-head detector system.

Fig. 4. The left image provides an example of an opposed dual-head imaging system while the system on the right is an example of a ring detector system.

PEM imaging is conducted with either a dual head or ring style gamma-ray detector. Both systems are designed to detect the coincident gamma rays which, are traveling approximately 180° from each other after the annihilation reaction (see figure 4). Unlike BSGI imaging devices, PEM devices do not use a collimator to help determine the location of each event. In PEM imaging, since there are two gamma-rays traveling 180° apart, the event location is calculated as a line of response between the location that each gamma-ray strikes the pair of opposed detectors. One advantage to PEM is that it does not have the same loss of resolution with distance that BSGI/MBI systems experience. As mentioned in the previous section, the mean free path of the 18F positron within the breast tissue is approximately 1 mm and commercially available PEM systems report an in-plane spatial resolution of about 2 mm. One limitation of the dual-head PEM detector design is that, due to the limited angle of acquisition, it has limited resolution in the Z-axis (depth). Ring detectors do not suffer from this limitation as they provide a 360° acquisition for reconstruction however there is currently no biopsy capability on the ring detector systems. A needle biopsy localization device was recently introduced for the opposed dual-head

Fig. 4. The left image provides an example of an opposed dual-head imaging system while

the system on the right is an example of a ring detector system.

Fig. 3. A typical 4-view BSGI study.

**3.1.3 Detectors for PEM** 

detector system.

One limitation to PEM image reconstruction is that the detector photon sensitivity is not linear across the field-of-view with lower sensitivity along the detector edges. This causes a higher level of noise to be present in the breast images, along the chest wall. Figure 5 provides a schematic representation of factors affecting the photon sensitivity in a PEM detector. The maximum angle of reconstruction (MAR) is a setting used in PEM software and it is defined as the maximum angle away from the detector normals (90 from the detector face) for which coincident gamma ray detections are included in tomographic image reconstruction. Larger MAR values yield greater overall photon sensitivity but with potential loss in spatial resolution due to the depth-of-interaction (DOI) blur. The blue lines in Figure 1 show the angular range over which gamma rays emitted from two points in the breast are accepted. Figure 1A shows a point near the nipple, and Figure 1B shows one near the chest. For events that occur in the center of the field-of-view (FOV), all of the events within the MAR are captured by the detector system. However a significant fraction of events occurring near the FOV edges go uncounted for because one of the paired gamma rays traveling outside the edges of the detector is not detected. As a recent study found, this loss of photon sensitivity along the edges limits the ability of the PEM system to detect lesions located near the chest wall (Rosen et al, 2005).

Fig. 5. A and B: a schematic example of the maximum angle of reconstruction near the center of the detector field-of-view and then near the chest wall respectively.

Fig. 6. a typical PEM study with multiple slice reconstruction for each projection.

The Role of Molecular Imaging Technologies in Breast Cancer Diagnosis and Management 189

Graph 1. The relative photon sensitivity of commercially available, breast-optimized

F-18 fluorodeoxy-D-glucose (FDG) was cleared by the US FDA in 2000 for a variety of uses including tumor localization. The total body radiation dose in FDG PET is 39 mrads per mCi injected activity (Table 5). According to the clinical literature, the typical FDG dose used for imaging with the standard whole body PET detectors ranges between is approximately 370 -

imaging systems compared to that of the standard gamma camera.

Table 4. Radiation dosimetry of Sestamibi.

**4.2 PEM radiation dose** 

740 MBq (10 - 20 mCi).

PEM detectors are tomographic imaging devices, an example image from the opposed detector system is provided in Figure 6. Note the noise level along the chest wall and the Z-resolution affect is expressed as a blurry, low intensity focus in the reconstruction planes outside of the plane the lesion is located in. In this particular case, it is most noticeable in the MLO projection images. There is noticeable residual blur in the area of the largest lesion in all of the projections, including those outside of the lesion.
