Boston University
OpenBU http://open.bu.edu
Theses & Dissertations Boston University Theses & Dissertations
2013
Does FDG-PET/MR offer a similar
diagnostic performance to
FDG-PET/CT, and therefore, offer a
comparable whole-body staging
examination in patients diagnosed
with non-Hodgkin's and Hodgkin's
lymphoma?
https://hdl.handle.net/2144/12049 Boston University
BOSTON UNIVERSITY SCHOOL OF MEDICINE
Thesis
DOES FDG-PET/MR OFFER A SIMILAR DIAGNOSTIC PERFORMANCE TO FDG-PET/CT, AND THEREFORE, OFFER A COMPARABLE WHOLE-BODY
STAGING EXAMINATION IN PATIENTS DIAGNOSED WITH NON-HODGKIN’S AND NON-HODGKIN’S LYMPHOMA?
by
WENDY LAURA ATKINSON
B.S., University of Illinois at Urbana-Champaign, 2007
Submitted in partial fulfillment of the requirements for the degree of
Master of Arts 2013
Approved by First Reader Hernan J. Jara, Ph.D. Professor of Radiology Second Reader Alexander R. Guimaraes, M.D., Ph.D. Assistant Professor of Radiology Harvard Medical School
Third Reader
Ciprian Catana, M.D., Ph.D. Assistant Professor of Radiology Harvard Medical School
DEDICATIONS
“When one door of happiness closes, another opens; But often we look so long at the closed door, That we do not see the one which has opened for us.”
– Helen Keller
ACKNOWLEDGEMENTS
AG & CC – I couldn’t have done this without you both. Thank you.
DOES FDG-PET/MR OFFER A SIMILAR DIAGNOSTIC PERFORMANCE TO FDG-PET/CT, AND THEREFORE, OFFER A COMPARABLE WHOLE-BODY
STAGING EXAMINATION IN PATIENTS DIAGNOSED WITH NON-HODGKIN’S AND NON-HODGKIN’S LYMPHOMA?
WENDY LAURA ATKINSON
Boston University School of Medicine, 2013 Major Professor: Hernan J. Jara, Ph.D., Professor of Radiology
ABSTRACT
Purpose
To evaluate the diagnostic performance of simultaneous FDG-PET/MR compared to FDG-PET/CT in non-Hodgkin’s and Hodgkin’s lymphoma.
Methods
This cross-sectional study included fifteen patients with a confirmed diagnosis of NHL or HL who had completed a clinical FDG-PET/CT and a same day research FDG-PET/MR. SUVmax for FDG-avid lesions were measured for each imaging modality, as well as ADC from FDG-PET/MR. Strength of correlation between variables was measured using the Spearman rank correlation coefficient (rs). The
overall radiation exposure dose was also calculated for a clinical FDG-PET/CT and compared to the radiation dose level remaining at time of FDG-PET/MR.
Results
Thirty-seven concordant FDG-avid lesions were identified on both PET/CT and PET/MR imaging. SUVmax from FDG-PET/MR versus FDG-PET/CT
demonstrated a strongly positive correlation (rs=0.84 (0.71, 0.92); p<0.0001). There was no correlation found between ADCmin and SUVmax from FDG-PET/MR (r=0.35(-0.07, 0.66); p=0.09). The overall radiation exposure from one FDG-PET/CT was 24.07±6.06mSv (range: 17.67-33.84mSv) compared to the decay-corrected radiation dose at FDG-PET/MR (2.87±0.92mSv (range: 1.86-5.90mSv)).
Conclusion
FDG-PET/MR offers a comparable whole body staging examination with an improved radiation safety profile in NHL and HL patients when based on the maximum standardized uptake value.
TABLE OF CONTENTS
Title i
Reader’s Approval Page ii
Dedication Page iii
Acknowledgements iv
Abstract v
Table of Contents vii
List of Tables viii
List of Figures ix List of Abbreviations x Introduction 1 Lymphoma 1 Imaging Modalities 4 Study Purpose 11 Methods 13 Subject Population 13 Study Design 14 Statistical Analysis 18 Results 19
Demographics & Clinical Parameters 19 PET/MR SUVmax Correlation to PET/CT SUVmax 21
ADCmin Correlation to PET/MR SUVmax 23
Radiation Dose Exposure 24
Discussion 26
Conclusion 34
References 36
LIST OF TABLES
Table Title Page
1 Patient Demographics and Clinical Parameters 20
2 SUVmax from PET/MR and PET/CT 21
LIST OF FIGURES
Figure Title Page
1 SUVmax Correlation Between PET/MR and PET/CT 22 2 Concordant and Disconcordant Lesions by PET and DWI 23 3
4
Correlation Analysis ADCmin of Compared to SUVmax Overall Radiation Exposure Dose by Radiation Source
24 26
LIST OF ABBREVIATIONS
AC attenuation correction
ADC apparent diffusion coefficient
ADCmin minimum apparent diffusion coefficient APD avalanche photodiode
AW-OSEM attenuation-weighted ordered subset expectation maximization BMI body mass ndex
CR complete remission CT computed tomography CTDIvol CT volume index
DLBCL Diffuse Large B Cell Lymphoma DLP dose length product
DWI diffuse-weighted imaging FDG [18F]Flurodeoxyglucose HL Hodgkin’s Lymphoma
MRI magnetic resonance imaging NHL Non-Hodgkin’s Lymphoma PET positron emission tomography PFS progression free survival PMT photomultiplier tube PR partial remission
ROI region of interest RS Reed-Sternberg SD standard deviation
SSEPI single shot echo planar imaging SUV standardized uptake value
INTRODUCTION
Lymphoma
Lymphoma is a type of cancer that occurs when certain immune cells called lymphocytes begin to grow or behave abnormally within the blood and is therefore termed a hematologic malignancy [1, 2]. The two major types of lymphocytes are B and T cells. Both play a role in recognizing and clearing the body of infectious organisms and abnormal cells. Lymphocytes circulate
throughout the body through a network of vessels containing a clear fluid called lymph. The lymphatic system includes the lymph nodes, thymus, tonsils, spleen, bone marrow, and the lymphoid tissue of the digestive system. Thousands of lymph nodes are scattered throughout the body, and each filters out microbial organisms from the lymph fluid. When lymphocytes begin to rapidly grow and divide, they often collect in one or more lymph nodes and ultimately form a mass called a tumor; however, it is possible for other non-lymph nodal sites to be
affected such as the liver, spleen, brain, skin, digestive system, and bone marrow [1]. Due to the lymphatic system, lymphomas may essentially spread to any area of the body[1, 2].
There are two major types of lymphoma: Hodgkin’s lymphoma (HL) and non-Hodgkin’s lymphoma (NHL). Lymphoma symptoms include, but are not limited to lymphadenopathy, night sweats, fever, weight loss, fatigue, and itching. While
both classifications of lymphoma often occur in the same locations within the body, they may present with the same symptoms, and also appear similar upon physical examination; they are differentiable at the microscopic level. [1-3].
Hodgkin’s lymphoma originates from a specific abnormal B cell lineage and includes 5 subtypes. At the microscopic level, Hodgkin cells are characterized by the presence of Reed-Sternberg (RS) cells which are readily identified by their large size and numerous and/or bi-lobed nuclei. HL typically follows a bimodal distribution (ages 15-34 years and above 55 years)[4]. Non-Hodgkin’s lymphoma includes over 60 different subtypes, and is essentially identified as any
lymphoma that does not involve RS cells. Due to this broad inclusion criteria, NHL is far more common in the United States, comprising over 85% of new lymphoma cases [5]. The incidence of NHL increases with age, with 50% of cases occurring in patients 65 years and older. Both NHL and HL are more prevalent in Caucasians and men in developed countries; however, there is no explanation for this observed prevalence. The overall 5-year period for survival is 69% for lymphoma; however, this figure can be as low as 10% for certain B and T cell NHL subtypes [5].
Similar to most types of cancer, therapeutic strategies and prognoses depend heavily on initial tumor staging [6]. Staging is a method to describe the extent and severity of cancer within the body. Prior to the start of the twenty-first century,
computed tomography (CT) was considered the integral method for evaluation and staging of lymphomas [7].
Imaging Modalities
Computed Tomography (CT)
CT uses computer-processed X-rays (ionizing radiation) to image the body in ‘slices’. Due to its high-contrast resolution, CT is able to detect density
differences in tissue that are as small as 1%. This is particularly advantageous for malignancies such as lymphoma, which exhibit increased cellularity (cellular density); although, CT is currently unable to distinguish between a viable tumor and a necrotic or fibrotic mass that remains after treatment [8]. This is an issue when trying to determine a patient’s response to treatment, or risk of relapse (cancer recurrence) after chemotherapy [8]. However, with the emergence of positron emission tomography (PET) with [18F]fluordeoxyglucose (FDG) in the late 1990s, oncologists now have an imaging tool that is able to overcome CT’s specificity limitation.
Positron Emission Tomography (PET)
PET is a form of nuclear/molecular imaging that is able to produce a three-dimensional tomographic image of functional processes within the body. When combined with FDG, PET is a critical tool for staging lymphoma patients [9]. FDG is a radioactive tracer with a chemical structure similar to glucose (sugar) that is
absorbed by “metabolically active” tissues such as malignant masses, and is unable to leave cells upon entering. During the PET scan, the F-18 isotopes decay and emit positrons (positively charged particle), which annihilate (collide) with electrons (negatively charged particles), releasing pairs of 511 keV photons (high energy particles) which are detected by the PET scanner. The more FDG that is taken up or absorbed by a specific tissue, the stronger the signal will be that is emitted. The signal intensity is indicative of the metabolic activity of the tumor. As a semi-quantitative measure, the PET signal is usually converted to standardized uptake values, which are obtained by normalizing the measured activity to the specific activity (injected dose over patient weight). One
measurement of uptake is the maximum standardized uptake value (SUVmax), which is routinely used for clinical purposes [10]. FDG-PET scans are
recommended before (initial staging), during, or after (restaging) treatment depending on the type and biological characteristics (fast growing vs. slow growing) of lymphoma [9]. Although PET is highly sensitive and specific for lymphoma (metabolically), it is unable to provide the precise anatomical information of CT [8].
Early interim FDG-PET (between chemotherapy cycles) has shown prognostic value in both NHL and HL[11-14]. In HL patients, early interim FDG-PET and SUVmax were identified as strong and independent predictors of progression free survival (PFS), and was superior to CT alone [14]. The two-year PFS in HL
patients was 100% for those with SUVmax < 3 g/ml [14]. In contrast, a positive early interim FDG-PET was highly predictive for progression in HL patients with advanced or extranodal disease. Similar results for predicting treatment failure and PFS have been reported for NHL patients [11, 12, 15, 16]. For primary staging, FDG-PET results, compared to CT alone, influenced disease staging in 15-20% of NHL patients, and altered clinical management in 5-10% [17, 18].
Hybrid PET/CT
In 2001, PET was combined with CT into a hybrid scanner called PET/CT [19]. This imaging modality combines the metabolic information provided by PET with the anatomical information provided by CT in order to localize the specific area of the body that exhibits FDG uptake. In addition, the combination of PET with CT has been able to address the issue of attenuation correction. Attenuation refers to signals (or coincidence events) that are lost (not recorded by PET detectors) due to body absorption or scattering. Attenuation correction is required for both qualitative and quantitative PET data analysis. Without attenuation correction, falsely intense or reduced FDG activity can be detected. PET-CT data is corrected for attenuation by converting the standard CT unitsa to linear
attenuation coefficients on a slice-by-slice basis. An attenuation map is created based on density differences which are obtained from the CT component of the scan. This, in turn, also allows for “co-registration” in which PET images are spatially correlated with CT images. In the last decade, PET-CT has replaced CT
as the standard of care for staging and assessing response to treatment in lymphoma [9, 20] and many other malignancies (e.g. lung, childhood
malignancies, colon cancer, and ovarian/cervical malignancies). Although PET-CT offers several advantages over the two separate imaging modalities, it is not without its own limitations.
Due to the need for numerous PET-CT examinations in a clinical oncology setting (initial staging, restaging, follow up), the amount of ionizing radiation received from each examination is a concern for patients with a good prognosis (such as lymphoma), and in particular, pediatric patients [21]. Numerous studies of atomic bomb survivors have shown a positive linear relationship between risk of solid cancer and radiation dose [22-25]. Furthermore, children are more radiosensitive (more susceptible to the harmful effects of ionizing radiation) than adults [25]. Based on the 2007 CT usage rates in the United States, it is
estimated that 1.5-2% of future cancers may be attributable to radiation from CT studies [26]. Perhaps even more alarming are the results from a 2004 study which reported that 53% of radiologists, 91% of ER physicians, and 97% of patients did not believe that CT scans increased lifetime cancer risks [27]. While there has been increased awareness of the potential heightened cancer risk from CT studies, results such as these behoove the medical community to seek out non-ionizing radiation diagnostic tools such as magnetic resonance imaging (MRI) where applicable.
Magnetic Resonance Imaging (MRI)
MRI is able to provide anatomical information similar to CT; however, MRI’s soft tissue differentiation ability is far superior to that of CT. In soft tissue organs and in areas with complex anatomy such as the brain, head and neck, and the pelvis, MRI allows for a more precise measurement of tumor location, size, and spread which all contribute to tumor staging [28]. MRI does not rely on ionizing radiation, but instead utilizes a powerful magnet and non-ionizing electromagnetic radiation in order to visualize structures within the body. The basic principle behind MRI is that human cells contain water molecules, the hydrogen atoms (protons) of which will align to the MRI’s magnetic field. This alignment will change when a source of electromagnetic radiation (in this case, a radio frequency) is applied, and the time it takes for the water molecules to return to their equilibrium (resting/normal) state is referred to as the relaxation time. Water molecules in different tissues of the body will have varying relaxation rates, and it is this difference (in addition to the actual proton density) that the MRI scanner can detect and use to create contrast between different structures and tissues. In addition, MRI is able to collect not only anatomical information, but also images of biophysical,
pathophysiological, or functional properties of tissues which are not possible on CT [28]. Lymphoma, in particular, is a type of cancer which stands to benefit from MR imaging.
Diffusion-Weighted MRI (DWI)
In recent years, the utility of MR imaging biomarkers in lymphoma patients has gained increased attention. MRI is capable of diffusion-weighted imaging (DWI-MRI) which is sensitive to the motion of water molecules in tissue [29]. Although DWI received notoriety initially as a diagnostic tool for the early assessment of cerebral ischemia, recently, DWI has been explored in tumors [30-34]. This movement of water is referred to as Brownian motion, which is an inherent, random motion of molecules. By comparing different diffusion images within the same area, the apparent diffusion coefficient (ADC) can be calculated which is a quantitative measurement of water movement [29]. As with most malignancies, lymphoma lesions exhibit high cellularity (cellular density), which can restrict diffusion (thereby producing an elevated signal on DWI), and results in lower ADC values than tissue with lower and/or normal cellularity. The lower an ADC value is in a given area (tumor), the more restricted the diffusion is, and the higher the signal will be on DWI. Several studies have shown that increased cellular density is correlated with reduced ADC values derived from diffusion MRI [30-32, 35]. A few studies have compared the diagnostic accuracy of DWI to PET/CT for lymphoma staging and response to treatment with promising results [10, 36, 37]. In addition, there has been data to suggest that a strong inverse correlation (e.g. one value increases while the other decreases) exists between ADC values and PET-SUV in lymphoma patients [34]. Based on these previous studies, the complementary nature of MRI and PET promises to be of great
clinical importance in the staging and response assessment of lymphoma patients.
Integrated PET/MR
One limitation of PET-CT is that PET and CT information is acquired sequentially (one after another) and not simultaneously (at the same time). This can result in potential image fusion issues as well as inaccurate image interpretation due to misalignment [28]. This is particularly relevant when imaging the chest and abdomen due to resultant motion from the respiratory and digestive systems. Simultaneously acquired data also permits a temporal relationship (matched in-time) to be established in which each data set is collected in exactly identical metabolic states. While MRI is able to overcome several of the limitations of CT, MRI as a separate imaging tool and thus, involving a separate scanning session, will still result in this same issue of misalignment and will also add increased burden on patients since they will have to undergo two separate imaging examinations (PET-CT and MRI).
The advent of an integrated whole-body PET/MR scanner in 2010 (Siemens Healthcare Biograph mMR) offers the possibility of obtaining PET’s metabolic (e.g. glucose metabolism, cell proliferation, etc.) information along with MRI’s anatomical and physiological (blood flow, vascular and tissue spaces, cellularity) information in one imaging session. Both types of information are currently used
in lymphoma staging [9]. This new integrated imaging modality also offers patients reduced radiation exposure compared to PET/CT by eliminating the CT component. The integration of PET and MR was made possible by the
development of PET detectors that are able to function within a strong magnetic field, but do not interfere with the MR system. The MR-compatible PET photon detectors contain avalanche photodiodes (APD) instead of the conventional photomultipliers tubes (PMT). Recent studies have demonstrated that the sensitivity of the PET component of the Biograph mMR is comparable to, or possibly even higher, than that of conventional PET-CT scanners [38] [39].
There are currently three PET/MR models on the market: Ingenuity TF (Philips), Discovery PET/CT+MRI (GE Healthcare), and the Biograph mMR (Siemens). The latter PET/MR model is a truly integrated unit, capable of performing
simultaneous PET and MRI [28]. The ability to acquire data simultaneously is of great clinical importance for numerous reasons. First, the temporally correlated measurements of early response to treatment can be assessed without a delay in time between two imaging modalities. This is particularly relevant for therapies with immediate effects (e.g. a drug with efficacy in hours rather than days) or studies involving continuously changing areas of the body (the digestive system). Second, simultaneously acquired data can also aid in MR-based motion
acquiring MR and PET data in one session provides added comfort and convenience to the patient by only requiring one imaging session.
Simultaneous PET/MR, albeit a promising imaging tool, is not without its own limitations and uncertainties [40]. Combined PET/MR faces unique challenges of attenuation correction of PET data. PET-CT data is corrected for attenuation by utilizing CT’s X-rays that are emitted to produce linear attenuation coefficients on a slice-by-slice basis. PET/MR, however, must rely on assigning attenuation coefficients by tissue class (based on MR data). This is challenging for many reasons, the largest of these being that MRI does not image cortical bone density. Instead, MR based attenuation correction is based on proton density (hydrogen atoms). Recent studies, however, have shown promising results for MR-based PET attenuation correction [41, 42]. While there is currently no
established clinical indication for PET/MR use [28], this newly integrated imaging modality holds great diagnostic promise in the field of oncology and in particular, lymphoma.
Study Purpose
The purpose of this research study was to explore the clinical utility of a relatively new dual imaging modality (PET/MR) compared to the current clinical standard of care (PET/CT), in patients with lymphoma, while also investigating the potential decrease in overall radiation exposure. Specifically, does FDG-PET/MR have a
similar diagnostic performance to FDG-PET/CT and therefore, offer a
comparable whole-body staging examination with an improved radiation safety profile in patients diagnosed with non-Hodgkin’s and Hodgkin’s lymphoma?
The diagnostic performance of FDG-PET/MR compared to FDG-PET/CT was evaluated based on the primary outcome of the maximum standardized uptake value (SUVmax) correlation between the two imaging modalities. The secondary outcomes of interest were the correlation of DWI-derived ADC values to FDG-PET/MR-derived SUVmax values, and the total received radiation dose per clinical PET/CT compared to decayed radiation dose at the time of the
FDG-PET/MR examination.
Based on current literature, there has not been a prior study evaluating the diagnostic utility of simultaneous FDG-PET/MR compared to FDG-PET/CT in lymphoma patients. Given the potential added value (MR-derived imaging biomarkers, soft tissue resolution, bone marrow lesion detection, and temporal correlation of PET and MR data) of this new imaging modality in lymphoma patients and the current lack of available data, there is a strong need to address the above research question. Prior studies have shown the potential utility of DWI-derived ADC values in lymphoma staging [10, 36, 37]; however, other studies have reported a less promising utility [43, 44]. None of these studies, however, has utilized simultaneous PET/MR due to its availability only in recent
years. Two studies have compared PET/MR performance to PET/CT with optimistic results for the former imaging modality, but neither has investigated this comparison in the lymphoma patient population. [39, 45].
MATERIALS AND METHODS
Subject Population
Informed consent was waived in this IRB-approved retrospective analysis of data previously collected. All original data were collected between October 2011 and February 2013. Fifty-one patients with known or suspected tumors gave informed consent for an IRB-approved prospective cohort study. Inclusion criteria in the original cohort included age between 18 and 80 years old, ability to give written informed consent, ability to lie comfortably flat, and received a clinical PET or PET/CT examination and a same-day research FDG-PET/MR (Biograph mMR, Siemens, Germany) examination. Exclusion criteria included pregnancy,
claustrophobia, and the presence of known contraindications for MRI including electrical or ferromagnetic implants.
The inclusion criteria for this retrospective analysis included a confirmed diagnosis of non-Hodgkin’s or Hodgkin’s lymphoma. The exclusion criteria included failure to complete both a clinical FDG-PET/CT and same day research FDG-PET/MR.
Study Design
This is a retrospective cross-sectional analysis of the data collected in the above prospective IRB-approved study.
PET/CT Protocol
All patients underwent a routine clinical PET-only or clinical PET/CT scan on a Siemens Biograph 64 (Siemens Healthcare, Germany). All patients fasted (except for water) for a 6 hour period before intravenous administration of the radiotracer [18F]fluordeoxyglucose (FDG), and blood glucose levels were checked to exclude hyperglycemia (>250 mg/dL) prior to FDG administration. The intravenously administered adult dose of FDG was based on body mass index (BMI): BMI <30 received 555MBq, BMI 31-44 received 740MBq, and BMI >45 received 925MBq. Imaging was performed 60 minutes after administration of FDG. Images were obtained from the base of the skull to the proximal one-third upper thigh.
PET/MRI Protocol
In all cases, imaging was performed with the Biograph mMR 3T scanner (Siemens Healthcare, Germany). Specially fabricated PET-compatible phased array coils (6-channel) were used to acquire the MR data. A 2-point Dixon 3D VIBE breath-hold T1 weighted sequence using the following parameters (iPAT factor 2, TR 3.6ms, TE1 1.225ms, TE2 2.45ms, matrix size 79x192, NEX=1, FOV
500mm, slice thickness 5.5mm, flip 10) was used to derive the attenuation
correction (AC) map. This manufacturer-provided method allows the identification of four tissue types (fat, soft tissue, lung, background). Being difficult to segment from the MR data, bone tissue is treated as soft tissue for AC purposes. Coronal whole-body STIR was performed (iPAT factor 3, TR 3382-5631ms, TE 81-87ms, matrix 186x384, NEX=1, FOV 450mm, slice thickness 5.5mm, TI 220-230ms) in order to provide anatomic correlation.
Whole-body DWI was performed using a single shot echo planar imaging technique (SSEPI) using the following parameters (iPAT factor 2, FOV 420, matrix 112x156, slice thickness 6mm, NEX 2, TE 68ms, TR 7800, TI 220ms (fat suppression), with b values including (0,50,800)).
PET data was acquired using shallow free breathing from the upper thighs cephalad. Four to five bed positions were required to cover the entire abdomen and pelvis with ~10 minute data acquisition per bed position. PET axial field of view was 25.8cm. PET data was reconstructed using a 3D attenuation-weighted ordered subset expectation maximization (AW-OSEM) algorithm, with 3 iterations and 21 subsets, zoom=1, and Gaussian smoothing of 4mm FWHM.
Maximum Standardized Uptake Value
Image analysis was performed using OsiriX (OsiriX®, Geneva, Switzerland) software with fusion software embedded. Region of interest (ROI) analysis was
performed by consensus with a staff radiologist and a nuclear medicine technician on FDG-PET data fused with either CT or MRI demonstrating FDG avidity defined as one standard deviation (SD) above liver or mediastinum. The maximum standardized uptake value (SUVmax) of the lesions was measured on both FDG-PET/MR and FDG-PET/CT.
Apparent Diffusion Coefficient (ADC)
ADC parametric maps were generated online on the Siemens mMR post-processing platform using a mono-exponential fit of the DWI raw data (b
=0,50,800 data). Image analysis was performed using OsiriX (OsiriX®, Geneva, Switzerland) software with fusion software embedded. Region of interest (ROI) analysis was performed by a staff radiologist using FDG-PET data fused with axial DWI data. Analysis included visibility of lesion on DWI data at b=50 and b=800 sec/mm2, in addition to quantitative measures of SUV as compared to ADC values.
Effective Radiation Dose
The radiation dose received from whole-body CT was based on methodology published elsewhere [46]. Briefly, present generation CT scanners generate a value called the dose-length product (DLP) which is a measure of the total
radiation delivered to a patient during a CT examination. The DLP was calculated from the length of the scan (cm) and the volume CT index (CTDIvol) which is an
index of the intensity of the radiation that the patient received: DLP = l (cm) x CTDIvol. Using the DLP value provided by the CT scanner as well as the number of slices and thickness per slice, an effective dose of the radiation received in milliSieverts (mSv) was obtained.
The effective radiation dose received from whole-body PET was based on methodology published elsewhere [47]. Briefly, the amount of activity of
radiotracer administered to a patient is a known value based on BMI. The degree of absorption of this activity varies depending on the organ referred to as the dose coefficient [48] and each organ was assigned a tissue weighting factor [49]. Effective radiation doses for each organ were calculated by multiplying the
radiotracer activity by both the dose coefficient and tissue weighting factor for that organ. The effective whole-body radiation dose was calculated by the sum function of all of the individual organ effective dose values.
Radioactivity at FDG-PET/MR
Using the known activity of radiotracer administered to each patient, the time administered, and the known half-life of with [18F]fluordeoxyglucose (109.77 min), the remaining activity at time of FDG-PET/MR scan (based on start time of PET acquisition) was calculated using GraphPad Radioactivity Calculator
(GraphPad Software, La Jolla, CA). This value was required in order to estimate the potential dose reduction from a unique PET/MR examination.
Statistical Analysis
Nonparametric correlation analyses were performed using the Spearman rank correlation coefficient (rs) was calculated to examine the correlation between the SUVmax derived from FDG-PET/MR and FDG-PET/CT, as well as the correlation between ADCmin and SUVmax derived from FDG-PET/MR using GraphPad Prism 6 (GraphPad Software, La Jolla, CA). Statistical significance was set at a p<0.05. Sample size was based on a power of 0.88 (beta of 0.12) and alpha of 0.05. In order to reject the null hypothesis (no relationship between two variables), a correlation coefficient of > 0.5 would need to be attained [50]. Correlation coefficient values of 0.5 – 1.00 indicate moderate to perfect strength of
correlation. The power was set at 0.88 since it is known that the Spearman Rank Correlation Coefficient has 91% of the power of a Pearson Correlation Coefficient (0.8/0.91=0.88) [51]. Based on the above alpha, beta, and r=0.5 in order to reject the null hypothesis, the sample size needed was 29 lesions. Lesions and not patients were chosen as the sample since each lesion produced unique values on the two imaging modalities and were located in various regions of the body, which provided the best overall assessment of the diagnostic performance
between FDG-PET/MR and FDG-PET/CT in lymphoma patients. For the purpose of this study, all FDG-avid lesions (n≥29) identified in eligible participants from the original cohort were included.
RESULTS
Demographics & Clinical Parameters
Of the 51 patients in the original data set, 17 patients (11 men and 6 women) had a confirmed diagnosis of non-Hodgkin’s or Hodgkin’s lymphoma. Two patients were excluded from this analysis due to incomplete FDG-PET/MR scans, leaving 15 NHL and HL patients for analysis in this retrospective study.
Patient demographics (sex, age, race), disease subtype, clinical exam (PET/CT or PET-only), number of FDG-avid lesions and number visualized on DWI are provided in Table 1 for all patients included in this analysis. Of the 15 subjects included in this analysis, 4 (26.67%) had a confirmed diagnosis of Hodgkin’s lymphoma and 11 (73.33%) had a confirmed diagnosis of non-Hodgkin’s
lymphoma (3 mantle cell lymphoma, 2 diffuse large B cell lymphoma (DLBCL), 2 marginal zone lymphoma, 2 follicular lymphoma, 1 mediastinal lymphoma, 1 post-transplantation lymphoproliferative disease). The mean age was
53.07±15.77 years (range: 30-80 years), and 60% (9/15) were male. Two subjects (13.33%) had received a clinical PET-only examination while the remaining thirteen subjects (86.67%) had received a clinical PET-CT
examination. For the purpose of analysis, PET-only and PET/CT data were analyzed as one cohort. Four subjects did not have any abnormal FDG uptake. Of the remaining 11 patients, 37 FDG-avid lesions were identified, 25 of those
exhibited abnormal diffusion on DWI. All FDG-avid lesions were identified on both PET/MR and PET/CT.
Table 1. Patient demographics and clinical parameters (DLBCL: Diffuse large B-cell lymphoma, HL: Hodgkin’s lymphoma, n/a: scan stopped early (excluded)
Patient no. Sex (M/F) Age (years)
Race Lymphoma Type No. of Lesions by PET No. of Lesions by DWI Lesion Location by PET 1 M 52 White Mediastinal unclassified
n/a n/a n/a 2 F 80 White DLBCL 9 8 Lung (6), abdomen (2), breast (1) 3 M 32 White HL 5 0 Lung (2), lymph nodes (2), bone (1) 4 M 48 White Blastoid mantle 3 2 Neck (2),
abdomen (1) 5 M 57 White Mantle cell 1 0 Neck (1) 6 F 32 White Mediastinal B-cell 0 -- -- 7 M 71 White Marginal zone 0 -- -- 8 M 32 Middle
Eastern
DLBCL 0 -- --
9 M 59 White Mediastinal B-cell n/a n/a n/a 10 M 30 White Mantle cell 0 -- -- 11 F 60 White HL 5 5 Peritoneal
(3), spleen (1), abdomen
(1) 12 M 59 White Follicular 7 5 Muscle (3),
subcutaneous (2), peritoneal (1), heart (1) 13 F 59 White Marginal zone 1 0 Rectum (1)
14 M 62 White HL 0 -- -- 15 M 71 White Post-transplantation lymphoproliferative disease 4 4 Liver (3), peritoneal (1) 16 F 53 White HL 1 0 Thyroid (1) 17 F 50 White Follicular 1 1 Heart (1)
PET/MR SUVmax correlation to PET/CT SUVmax
Thirty-seven FDG-avid lesions were analyzed. The mean SUVmax from PET/CT was on average lower than that of PET/MR with mean 4.59±2.72 g/ml (range: 1.1-10.9 g/ml) and the mean PET/MR SUVmax being 8.50±4.57 g/ml (range: 2.8-19.4; Table 2). An upward trend (before-after plot) of PET/MR SUVmax values compared to subject-matched PET/CT SUVmax values is demonstrated in Figure 1a. Spearman rank correlation analysis comparing FDG-PET/MR SUVmax to FDG-PET/CT SUVmax showed a moderate to strong correlation, which was statistically significant (rs=0.84 (0.71, 0.92), p<0.0001; Figure 1b).
Table 2: SUVmax from PET/MR and PET/CT (n=37) Imaging Modality SUVmax Mean
(g/ml)
SD Range
PET/MR 8.50 4.57 2.8, 19.4
(a)
(b)
Figure 1. SUVmax correlation between PET/MR and PET/CT. (a) Before-after plot
demonstrating upward trend of maximum SUV from PET/CT and matched PET/MR; (b) Correlation analysis of maximum SUV of PET/MR versus maximum SUV of FDG-PET/CT demonstrating strong correlation (rs=0.84; p<0.0001).
ADCmin correlation to PET/MR SUVmax
Twenty-five of the thirty-seven FDG-avid lesions (green arrow) also exhibited restricted diffusion on DWI (Figure 2a, green arrow), whereas twelve of the FDG-avid lesions did not and were considered “missed” (Figure 2b, yellow arrow).
The mean ADCmin was 679.88±326.58mm2s-1 (range: 142-1472mm2s-1), and the mean SUVmax from PET/MR was 8.30±4.79 g/ml (range: 2.8-19.4). The mean
Figure 2. Concordant and disconcordant lesions by PET and DWI. (a) Breast lesion with FDG avidity on PET/MR (green arrow); (b) Breast lesion with concordant restricted diffusion on DWI (green arrow); (c) Two mediastinal lesions exhibiting equal FDG avidity on PET/MR (green arrows); (d) Right mediastinal lesion exhibiting concordant restricted diffusion on DWI (green arrow). Left mediastinal lesion exhibiting decreased restricted diffusion (decreased signal on DWI) compared to right lesion (yellow arrow).
PET/MR SUVmax was not the same as the mean used to calculate the Spearman correlation above since only 25 (of 37) FDG-avid lesions were visualized on DWI. Spearman rank correlation analysis compared to FDG-PET/MR showed a weak to moderate correlation that was not statistically significant (r=0.35(-0.07, 0.66), p=0.09; Figure 3).
Radiation dose exposure
Thirteen subjects with completed clinical PET/CT examinations were included in the total FDG-PET/CT radiation dose analysis. Two of the patients in this study PET only examinations, and while a low dose of CT radiation was administered for attenuation correction purposes, the DLP values were not recorded per
Figure 3. Correlation analysis of ADCmin compared to SUVmax. Maximum SUV
from FDG-PET/MR demonstrated a weakly positive correlation coefficient to ADCmin that was not statistically significant (r=0.35, p=0.09).
standard practice at this institution. The mean total radiation dose from PET/CT was 24.07±6.06mSv (range: 17.67-33.84mSv; Table 3, Figure 4). Of this total, the dose from CT accounted for approximately 50% with a mean CT dose of 11.93±6.55mSv (range: 8.83-21.13mSv) and the mean FDG-PET dose was 10.60±1.76mSv (range: 9.16-13.83mSv). In comparison, the mean radiation dose at time of FDG-PET/MR scan, corrected for decay over time, was
2.87±0.092mSv (range: 1.54-5.00mSv) for the fifteen subjects who completed an FDG-PET/MR scan.
Table 3: Ionizing radiation dose exposure (*: Decay-corrected radiation dose at time of FDG-PET/MR; **: Two patients had PET-only exams)
Radiation Source N** Dose Mean (mSv) SD Range CT 13 11.93 6.55 8.83, 21.13 FDG-PET 15 10.60 1.76 9.16, 13.83 Total FDG-PET/CT 13 24.07 6.06 17.67, 33.84 *FDG-PET/MR 15 3.41 1.12 1.86, 5.90
DISCUSSION
In the present study, the diagnostic performance of FDG-PET/MR in lymphoma patients was compared to FDG-PET/CT using the maximum standardized uptake value and the apparent diffusion coefficient. In addition, the radiation dose profile between the two imaging modalities was compared. Based on current
knowledge, no other study has compared FDG-PET/MR to FDG-PET/CT on the basis of SUVmax correlation, ADC to SUVmax correlation, and radiation dose in patients with non-Hodgkin's and Hodgkin’s lymphoma.
Qualitatively, there was no discordance between FDG avid lesions comparing FDG PET/CT and FDG PET/MR. This is critical when comparing the diagnostic performance and accuracy between these two technologies, as it conveys that
Figure 4. Overall radiation exposure dose by radiation source. (*): FDG level corrected for decay at start of PET acquisition.
no changes in either staging or treatment response were observed, and therefore no change in management would have been conveyed. Furthermore, the
primary outcome in this study was the correlation of SUVmax derived from PET/MR to SUVmax from clinical PET/CT. The Spearman correlation coefficient was rs=0.85 (P<0.0001) which indicates a strongly positive correlation between the SUVmax derived from the two imaging modalities. This correlation coefficient is slightly lower than results reported by Drzezga et al. and McDermott et al. [39, 52] who reported a very high correlation of rs=0.93 and rs =0.92, respectively, between mean SUVs in suspicious lesions in PET/MR and PET/CT. Interestingly, in that study, the SUV mean values for lesions on PET/CT were higher than those obtained for PET/MR (9.4±18.4% PET/CT > PET/MR). In contrast, the present study found higher on average SUV mean values for PET/MR compared to PET/CT (8.50±4.89 g/ml compared to 4.59±2.89 g/ml). Since there has only been one prior study directly comparing SUVmax values in PET/CT and PET/MR as mentioned above, there is no consensus on whether FDG uptake on PET/MR is typically higher or lower than on PET/CT. The amount of time elapsed from time of FDG administration to time of imaging has been shown to have a strong effect on the SUV measured [53]. A study by Hamberg et al. reported that SUV levels did not plateau for up to five hours post FDG administration in lung cancer patients [54]. While the same study has not been conducted in lymphoma
patients, Hamberg et al.’s results support the higher PET/MR-derived SUV values reported in the present study Moreover, the absolute value of SUVmax is
not as clinically important as is the relative change from baseline or prior FDG-PET.
This study also evaluated the relationship between DWI-derived ADC values and SUVmax. In some previous studies, ADC values were reported to correlate
inversely with SUVmax in both lymphoma and other malignancies such as cervical cancer [30, 32] [55]. Conversely, Wu et al. [56] showed that there was no
correlation between ADC values and SUVmax in non-Hodgkin’s lymphoma. All of these studies, however, relied on diffusion-MRI and PET data obtained from two different imaging modalities, and hence, two different imaging sessions. The results presented in this study do not support a strong correlation between ADC and SUVmax. Analysis of these two measurements resulted in a Spearman rank coefficient of rs=0.35 (p<0.09), which was not statistically significant. It is possible that in the quest to overcome the numerous challenges of combining MR and PET imaging, the absolute value of the apparent diffusion coefficient or the maximum standardized uptake value may be skewed in the process. As reported above, the observed SUVmax values from simultaneous PET/MR were higher on average than those reported by Drzezga et al. [39]. This relative difference could potentially reduce the presence of an inverse correlation between two variables, and could explain the lack of strong correlation observed.
It should be noted that not all FDG-avid lesions exhibited abnormal diffusion on DWI and as a result, ADC values could not be obtained for those lesions. The ADC results presented in this study were only abnormal in 25/37 identified lesions. Of the twelve “missed” lesions, four were located in the chest (5/12 in neck, 1/12 in abdomen, 1/12 in muscle, 1/12 in thyroid). This variation in DWI is consistent with previous research in which signal loss and motion artifact
occurred in the chest region due to respiration [57]. While this study did not support a correlation between ADC and SUVmax, it is important to acknowledge the potential diagnostic utility of these factors individually when staging and treating lymphoma patients.
It should also be noted that these lymphoma patients were at various stages of their treatment. None of the patients were at their baseline (pre-treatment) measurement. Therefore, this lack of correlation may be indicative and supportive that their measures were reflective of two different physiologic qualities (i.e. metabolism (FDG) and cellular density (ADC)). This may support their complementary nature and reflect another means of stratification of this patient population. Furthermore, since DWI is truly non-invasive, it comes at no risk for the patient.
The overall radiation dose received from one clinical FDG-PET/CT was
This amount of radiation exposure is approximately equal to 6-7 years of natural background radiation [58] and at the present time, poses an unknown cancer risk. While this is not particularly relevant in many oncologic subtypes, given the average lymphoma patient’s age and lifetime mortality risk from their cancer, it is particularly relevant in this sub-population, many of whom are younger (even children), and many of whom are deemed “cured” after systemic chemotherapy. For those still with FDG avid disease, the International Harmonization Project criteria recommends baseline and post-therapeutic intervention PET/CT scans for NHL and HL patients [9]. Depending on the response to treatment, lymphoma patients are likely to have several PET/CT scans per year until partial remission (PR) and/or complete remission (CR) is achieved. This number, and as a result, a patient’s radiation exposure will increase if relapse occurs. While 80% of relapse cases can be identified clinically, subsequent imaging is required to determine restaging [9, 59]. Since radiation exposure is cumulative, a patient may be exposed to the equivalent of up to twenty years of background radiation in a single year due to numerous PET/CT examinations.
In comparison, the remaining level of radioactivity (decay due to time delay between scans) at the time of the PET/MR scan had a mean of 3.28±0.92mSv. While this measurement is an estimate of the actual radiotracer dose that is currently used clinically for FDG-PET, it is suggestive of a potential reduction in FDG radiation exposure for oncology, in particular children and lymphoma,
patients in the future. It is possible that the high sensitivity of PET/MR can ultimately lead to a lower radiation dose to a level somewhere between the decayed-corrected FDG mean reported in this study and the mean clinical FDG-PET dose administered. This idea is supported by the fact that all FDG-avid lesions on PET/CT were identified on PET/MR despite the lapse in time and radiation decay, in addition to the strong correlation of PET-MR SUVmax to PET-CT SUVmax; however, the validity of this statement is out of the scope of this study and it remains to be seen if and by how much FDG-PET doses can be reduced while maintaining high sensitivity and specificity. Since the CT radiation dose comprised approximately 50% of the total FDG-PET/CT in this study, removing this source of ionizing radiation while leaving the FDG-PET dose the same would still result in a significant reduction in overall radiation exposure for the patient. Furthermore, it should be noted that this Institution uses CT radiation levels that are 30% to 95% less than reference levels used by the National Council on Radiation Protection [60]. This supports the possibility of an even larger reduction (more than 50%) in radiation exposure by utilizing PET/MR in place of PET/CT for lymphoma staging and response to treatment across the United States and abroad. Due to the cumulative nature of radiation exposure and the risk of cancer induced from radiation [47], the medical community should carefully weigh the risk-benefit ratio when ordering diagnostic examinations involving ionizing radiation such as PET/CT. This is particularly relevant as
further studies evaluate the diagnostic utility of simultaneous PET/MR and the technology gains further and wider acceptance.
A major strength of this study is that it was able to explore the association of a quantitative metabolic measurement from truly simultaneous PET/MR with
PET/CT. This is only the second study to explore this potential relationship, and it is the only study to focus on lymphoma patients. Furthermore, there were no FDG-avid lesions detected on PET/CT that were missed on PET/MR, and strong correlation coefficient reported here is similar to the one reported by Drzezga et al.[39].
In addition, the demographics demonstrated in this study are representative of lymphoma patients in the United States. Compared to 85% of lymphoma patients having the non-Hodgkin’s subtype, this study included 73.33% NHL patients, with mean age of 53.07±15.77 years (accounting for the bimodal distribution of HL patients), predominantly white, and 60% (9/15) were male which equals the 1.5:1 incidence ratio of men to women.
A notable limitation of this study is the number of available lymphoma patients who had received a clinical FDG-PET/CT and research FDG-PET/MR.
Specifically, the number of FDG-avid lesions was variable for each patient, ranging from no abnormal FDG uptake to up to ten identifiable lesions. While it is
difficult to speculate how increasing the n would change the results of this study, a larger study group would allow for stratification by lesion location (chest, neck, pelvis) or lymphoma subtype which could provide added value for PET/MR clinical indications. Furthermore, while a statistical adjustment was not made for the number of lesions per patient, a future study might want to address this point to remove this potential confounder. Another limitation is that the lymphoma patients included in this analysis were at various stages of their treatment, which could account for those patients with no FDG-avid lesions (if PET/CT was post-treatment), and be a potential source of poor correlation between SUV and ADC. The radiologist who determined the ROIs used to calculate the SUVmax and ADC values was not blinded to this study; however, designing such a blinded study poses a great challenge as the anatomic images from PET/CT and PET/MR images are visually different despite the similarities between the PET images. Another potential limitation is the MR-based attenuation correction which may vary in accuracy based upon the body region, and is an area for further
investigation. Lastly, a potential limitation in this study is the inclusion of
standardized uptake values from PET-only examinations; however, a low-dose CT is used to correct for attenuation and is equivalent to the attenuation
correction that occurs in PET/CT at this Institution.
Integrated PET/MR is a relatively new imaging modality with numerous potential applications in the field of oncology and areas yet to be identified. The limited
numbers of studies investigating the diagnostic utility of such an integrated system have been promising thus far, and further studies are warranted. Within the indication of NHL and HL, a larger prospective cohort study aimed at
validating the diagnostic utility of PET/MR compared to PET/CT should be undertaken. Specifically, the sensitivity and specificity of FDG-avid lesion detection as well as the complementary nature of diffusion-MR should be evaluated further. Lymphoma patients stand to benefit directly from the availability of clinical PET/MR scans compared to PET/CT due to the reduced levels of ionizing radiation, and thus, may benefit from a lowered lifetime risk of cancer.
CONCLUSION
The purpose of this research study was to explore the clinical utility of a relatively new dual imaging modality (FDG-PET/MR) compared to the current clinical standard of care (FDG-PET/CT), while also investigating the potential decrease in overall radiation exposure. The primary outcome in this study was the
maximum standardized uptake value (SUVmax) correlation between the two imaging modalities. The secondary outcomes of interest were the correlation of DWI-derived ADC values to FDG-PET/MR-derived SUVmax values, and the total received radiation dose per clinical FDG-PET/CT compared to decayed radiation dose at FDG-PET/MR. The Spearman rank correlation coefficient was used to
determine the correlation between SUVmax derived from PET/MR and FDG-PET/CT, resulting in a ranked correlation coefficient of rs=0.84 (0.71, 0.92), p<0.0001. The correlation of ADCmin compared to SUVmax derived from FDG-PET/MR was also calculated using the Spearman rank correlation coefficient, resulting in a ranked correlation coefficient of rs=0.35 (-0.07, 0.66), p=0.09.
The results of this study support the conclusion that the diagnostic performance of FDG-PET/MR compared to FDG-PET/CT as measured by SUVmax is similar between the two imaging modalities, and therefore, offers a comparable whole body staging examination. However, the results of this study do not support the conclusion that ADCmin is correlated to SUVmax in an integrated PET/MR scanner. Fifty percent of the high levels of ionizing radiation received from combined PET/CT are attributable to radiation from CT alone, and eliminating this component in an integrated PET/MR system supports the conclusion that PET/MR use can lead to a decrease in a patient’s overall radiation exposure.
REFERENCES
1. Society, A.C., Non-Hodgkin Lymphoma, 2013, American Cancer Society: Atlanta, GA.
2. Society, A.C., Hodgkin Lymphoma, 2013, American Cancer Society: Atlanta, GA.
3. Society, A.C., Cancer Facts & Figures 2012, 2013, American Cancer Society: Atlanta, GA.
4. Medeiros, L.J. and T.C. Greiner, Hodgkin's disease. Cancer, 1995. 75(1 Suppl): p. 357-69.
5. Howlader, N., et al., SEER Cancer Statistics Review, 1975-2009 (Vintage
2009 Populations), 2011, National Cancer Institute: Bethesda, MD.
6. Lin, C., et al., Whole body MRI and PET/CT in haematological
malignancies. Cancer imaging : the official publication of the International
Cancer Imaging Society, 2007. 7 Spec No A: p. S88-93.
7. Lister, T.A., et al., Report of a committee convened to discuss the
evaluation and staging of patients with Hodgkin's disease: Cotswolds meeting. Journal of clinical oncology : official journal of the American
Society of Clinical Oncology, 1989. 7(11): p. 1630-6.
8. Stumpe, K.D., et al., Whole-body positron emission tomography using
fluorodeoxyglucose for staging of lymphoma: effectiveness and comparison with computed tomography. European journal of nuclear
medicine, 1998. 25(7): p. 721-8.
9. Cheson, B.D., et al., Revised response criteria for malignant lymphoma. Journal of clinical oncology : official journal of the American Society of Clinical Oncology, 2007. 25(5): p. 579-86.
10. Abdulqadhr, G., et al., Whole-body diffusion-weighted imaging compared
with FDG-PET/CT in staging of lymphoma patients. Acta radiologica,
2011. 52(2): p. 173-80.
11. Jerusalem, G., et al., Persistent tumor 18F-FDG uptake after a few cycles
of polychemotherapy is predictive of treatment failure in non-Hodgkin's lymphoma. Haematologica, 2000. 85(6): p. 613-8.
12. Mikhaeel, N.G., et al., 18-FDG-PET as a prognostic indicator in the
treatment of aggressive Non-Hodgkin's Lymphoma-comparison with CT.
Leukemia & lymphoma, 2000. 39(5-6): p. 543-53.
13. Spaepen, K., et al., Early restaging positron emission tomography with (
18)F-fluorodeoxyglucose predicts outcome in patients with aggressive non-Hodgkin's lymphoma. Annals of oncology : official journal of the
European Society for Medical Oncology / ESMO, 2002. 13(9): p. 1356-63.
14. Hutchings, M., et al., FDG-PET after two cycles of chemotherapy predicts
treatment failure and progression-free survival in Hodgkin lymphoma.
Blood, 2006. 107(1): p. 52-9.
15. Brepoels, L., et al., Hodgkin lymphoma: Response assessment by revised
International Workshop Criteria. Leukemia & lymphoma, 2007. 48(8): p.
1539-47.
16. Hoekstra, O.S., et al., Early treatment response in malignant lymphoma,
as determined by planar fluorine-18-fluorodeoxyglucose scintigraphy.
Journal of nuclear medicine : official publication, Society of Nuclear Medicine, 1993. 34(10): p. 1706-10.
17. Moog, F., et al., Extranodal malignant lymphoma: detection with FDG PET
versus CT. Radiology, 1998. 206(2): p. 475-81.
18. Shelly, M.J., et al., 18-Fluorodeoxyglucose Positron Emission
Tomography/Computed Tomography in the Management of Aggressive Non-Hodgkin's B-Cell Lymphoma. ISRN hematology, 2012. 2012: p.
19. Beyer, T., et al., A combined PET/CT scanner for clinical oncology. Journal of nuclear medicine : official publication, Society of Nuclear Medicine, 2000. 41(8): p. 1369-79.
20. Delbeke, D., et al., Expert opinions on positron emission tomography and
computed tomography imaging in lymphoma. The oncologist, 2009. 14
Suppl 2: p. 30-40.
21. Pearce, M.S., et al., Radiation exposure from CT scans in childhood and
subsequent risk of leukaemia and brain tumours: a retrospective cohort study. Lancet, 2012. 380(9840): p. 499-505.
22. Pierce, D.A. and D.L. Preston, Radiation-related cancer risks at low doses
among atomic bomb survivors. Radiation research, 2000. 154(2): p.
178-86.
23. Preston, D.L., et al., Studies of mortality of atomic bomb survivors. Report
13: Solid cancer and noncancer disease mortality: 1950-1997. Radiation
research, 2003. 160(4): p. 381-407.
24. Preston, D.L., et al., Effect of recent changes in atomic bomb survivor
dosimetry on cancer mortality risk estimates. Radiation research, 2004.
162(4): p. 377-89.
25. Preston, D.L., et al., Solid cancer incidence in atomic bomb survivors:
1958-1998. Radiation research, 2007. 168(1): p. 1-64.
26. Brenner, D.J. and E.J. Hall, Computed tomography--an increasing source
of radiation exposure. The New England journal of medicine, 2007.
357(22): p. 2277-84.
27. Lee, C.I., et al., Diagnostic CT scans: assessment of patient, physician,
and radiologist awareness of radiation dose and possible risks. Radiology,
28. Balyasnikova, S., et al., PET/MR in oncology: an introduction with focus
on MR and future perspectives for hybrid imaging. American journal of
nuclear medicine and molecular imaging, 2012. 2(4): p. 458-74.
29. Koh, D.M. and D.J. Collins, Diffusion-weighted MRI in the body:
applications and challenges in oncology. AJR. American journal of
roentgenology, 2007. 188(6): p. 1622-35.
30. Sumi, M., Y. Ichikawa, and T. Nakamura, Diagnostic ability of apparent
diffusion coefficients for lymphomas and carcinomas in the pharynx.
European radiology, 2007. 17(10): p. 2631-7.
31. Toh, C.H., et al., Primary cerebral lymphoma and glioblastoma multiforme:
differences in diffusion characteristics evaluated with diffusion tensor imaging. AJNR. American journal of neuroradiology, 2008. 29(3): p. 471-5.
32. Nakayama, T., et al., Usefulness of the calculated apparent diffusion
coefficient value in the differential diagnosis of retroperitoneal masses.
Journal of magnetic resonance imaging : JMRI, 2004. 20(4): p. 735-42.
33. Padhani, A.R., et al., Diffusion-weighted magnetic resonance imaging as a
cancer biomarker: consensus and recommendations. Neoplasia, 2009.
11(2): p. 102-25.
34. Punwani, S., et al., Quantitative diffusion weighted MRI: a functional
biomarker of nodal disease in Hodgkin lymphoma? Cancer biomarkers :
section A of Disease markers, 2010. 7(4): p. 249-59.
35. Barajas, R.F., Jr., et al., Diffusion-weighted MR imaging derived apparent
diffusion coefficient is predictive of clinical outcome in primary central nervous system lymphoma. AJNR. American journal of neuroradiology,
2010. 31(1): p. 60-6.
36. Punwani, S., et al., Diffusion-weighted MRI of lymphoma: prognostic utility
and implications for PET/MRI? European journal of nuclear medicine and
37. Gu, J., et al., Whole-body diffusion-weighted imaging: the added value to
whole-body MRI at initial diagnosis of lymphoma. AJR. American journal
of roentgenology, 2011. 197(3): p. W384-91.
38. Delso, G., et al., Performance measurements of the Siemens mMR
integrated whole-body PET/MR scanner. Journal of nuclear medicine :
official publication, Society of Nuclear Medicine, 2011. 52(12): p. 1914-22.
39. Drzezga, A., et al., First clinical experience with integrated whole-body
PET/MR: comparison to PET/CT in patients with oncologic diagnoses.
Journal of nuclear medicine : official publication, Society of Nuclear Medicine, 2012. 53(6): p. 845-55.
40. Yankeelov, T.E., et al., Simultaneous PET-MRI in oncology: a solution
looking for a problem? Magnetic resonance imaging, 2012. 30(9): p.
1342-56.
41. Catana, C., et al., Toward implementing an MRI-based PET
attenuation-correction method for neurologic studies on the MR-PET brain prototype.
Journal of nuclear medicine : official publication, Society of Nuclear Medicine, 2010. 51(9): p. 1431-8.
42. Marshall, H.R., et al., A comparison of MR-based attenuation correction in
PET versus SPECT. Physics in medicine and biology, 2011. 56(14): p.
4613-29.
43. Wu, X., et al., ADC measurements in diffuse large B-cell lymphoma and
follicular lymphoma: a DWI and cellularity study. European journal of
radiology, 2013. 82(4): p. e158-64.
44. van Ufford, H.M., et al., Newly diagnosed lymphoma: initial results with
whole-body T1-weighted, STIR, and diffusion-weighted MRI compared with 18F-FDG PET/CT. AJR. American journal of roentgenology, 2011.
196(3): p. 662-9.
45. Wiesmuller, M., et al., Comparison of lesion detection and quantitation of
PET/MR hybrid scanner and PET from PET/CT. European journal of
nuclear medicine and molecular imaging, 2013. 40(1): p. 12-21.
46. Huda, W., K.M. Ogden, and M.R. Khorasani, Converting dose-length
product to effective dose at CT. Radiology, 2008. 248(3): p. 995-1003.
47. Huang, B., M.W. Law, and P.L. Khong, Whole-body PET/CT scanning:
estimation of radiation dose and cancer risk. Radiology, 2009. 251(1): p.
166-74.
48. 1990 Recommendations of the International Commission on Radiological
Protection. Annals of the ICRP, 1991. 21(1-3): p. 1-201.
49. The 2007 Recommendations of the International Commission on
Radiological Protection. ICRP publication 103. Annals of the ICRP, 2007.
37(2-4): p. 1-332.
50. Zou, K.H., K. Tuncali, and S.G. Silverman, Correlation and simple linear
regression. Radiology, 2003. 227(3): p. 617-22.
51. Schatzinger, R.A. and J.F. Jordan, Reservoir characterization : recent
advances. AAPG memoir1999, Tulsa, Okla.: American Association of
Petroleum Geologists. ix, 404 p.
52. McDermott, S., Blake, M, Sahani, D, Catana, C, Rosen, B, Guimaraes, A., Maximum SUV: Do PET/MR and PET/CT Differ? Our Experience., in
Radiological Society of North America 2012 Scientific Assembly and Annual Meeting2012: Chicago, IL. .
53. Keyes, J.W., Jr., SUV: standard uptake or silly useless value? Journal of nuclear medicine : official publication, Society of Nuclear Medicine, 1995.
36(10): p. 1836-9.
54. Hamberg, L.M., et al., The dose uptake ratio as an index of glucose
metabolism: useful parameter or oversimplification? Journal of nuclear
medicine : official publication, Society of Nuclear Medicine, 1994. 35(8): p. 1308-12.
55. Ho, K.C., et al., Correlation of apparent diffusion coefficients measured by
3T diffusion-weighted MRI and SUV from FDG PET/CT in primary cervical cancer. European journal of nuclear medicine and molecular imaging,
2009. 36(2): p. 200-8.
56. Wu, X., et al., No correlation between glucose metabolism and apparent
diffusion coefficient in diffuse large B-cell lymphoma: a PET/CT and DW-MRI study. European journal of radiology, 2011. 79(2): p. e117-21.
57. Kwee, T.C., et al., Diffusion-weighted whole-body imaging with
background body signal suppression (DWIBS): features and potential applications in oncology. European radiology, 2008. 18(9): p. 1937-52.
58. Measurements, N.C.o.R.P.a., Ionizing Radiation Expsoure of the
Popuation of the United States, 2009: Bethesda, MD.
59. Weeks, J.C., et al., Value of follow-up procedures in patients with
large-cell lymphoma who achieve a complete remission. Journal of clinical
oncology : official journal of the American Society of Clinical Oncology, 1991. 9(7): p. 1196-203.
VITA
