Title Radiation dose and cancer risk of cardiac CT scan and PET CT

Document Sample
Title Radiation dose and cancer risk of cardiac CT scan and PET CT Powered By Docstoc
					             Radiation dose and cancer risk of cardiac CT scan and
   Title     PET-CT scan


Author(s)    Huang, Bingsheng; žÃp³SG



 Citation



Issue Date   2009



  URL        http://hdl.handle.net/10722/54686



  Rights     unrestricted
RADIATION DOSE AND CANCER RISK
                 OF
CARDIAC CT SCAN AND PET-CT SCAN




          HUANG BINGSHENG


             MPHIL THESIS




      THE UNIVERSITY OF HONG KONG
                 2009
RADIATION DOSE AND CANCER RISK
                      OF
CARDIAC CT SCAN AND PET-CT SCAN

                  Submitted by



             Huang Bingsheng

          B.Eng. (Tsinghua University)
          M.Eng (Tsinghua University)




      for the degree of Master of Philosophy
         at The University of Hong Kong


                in February 2009
ABSTRACT




                 ABSTRACT OF THESIS ENTITLED




     RADIATION DOSE AND CANCER RISK OF
      CARDIAC CT SCAN AND PET-CT SCAN

                                     Submitted by



                             Huang Bingsheng

 For the degree of Master of Philosophy at the University of Hong Kong in February 2009


In our study, we aimed to measure and calculate the radiation doses to organs
resulting from (1) 64-slice cardiac computed tomography (CT) scan in pediatric
patients, (2) 64-slice coronary CT angiography (CTA) in adult patients, (3) whole
body positron emission tomography-computed tomography (PET-CT) scan in adult
patients. We also aimed to estimate the cancer risks caused by these radiation doses,
for the United States (US) and the Hong Kong (HK) population.


The organ doses from CT scan were measured using humanoid phantoms and
thermoluminescent dosimeters (TLD), and were also simulated using dosimetry
calculator (ImPACT). Dose coefficients recommended in International Commission
on Radiological Protection (ICRP) publication 80 were used to calculate the organ
doses from PET scan. Effective doses were calculated from organ doses using the
tissue weighting factors from ICRP publication 103. Tables of lifetime attributable
risk (LAR) of cancer incidence for the US and HK population were updated or set up
by using the method introduced in BEIR VII report of National Research Council,
and were used for estimating the LAR associated with these doses applying linear
extrapolation.


The effective doses of pediatric cardiac CT scan was 11.88 mSv~16.56 mSv and
11.90 mSv~16.57 mSv for boys and girls respectively. The effective doses of adult

                                           I
ABSTRACT


coronary CTA were 22.85 mSv~27.33 mSv and 20.76 mSv~23.32 mSv for males and
females respectively. For the combined PET-CT scan, the total effective doses were
13.65 mSv~32.18 mSv for males and 13.45 mSv~31.91 mSv for females.


The LARs induced by pediatric cardiac CT scan were 0.22%~0.33% and 0.61%~
0.85% for 5-year-old HK boys and girls respectively; and were 0.14%~0.20% and
0.43%~0.60% for 5-year-old US boys and girls respectively. The LAR induced by a
single CTA was 0.19%~0.23% and 0.59%~0.69% for 20-year-old HK males and
females respectively; and were 0.14%~0.17% and 0.46%~0.55% for 20-year-old US
males and females respectively. For whole body PET-CT scan, the LAR induced was
0.17%~0.39% and 0.27%~0.62% for 20-year-old HK males and females respectively;
and were 0.16%~0.32% and 0.23%~0.55% for 20-year-old US males and females
respectively.


LARs for women are higher than men across all ages and scan types. This is mainly
because girls’ breasts receive a high dose and carries a high risk of developing cancer.
It is also found that the risks are higher in the HK population than in the US
population. This is due to the different cancer statistics and life span data for the HK
and US populations. The risks were higher in the younger population than in the
older population, and especially high in children, because children have a longer life
expectancy and are more sensitive to radiation.


It is concluded that the cardiac CT and PET-CT examinations should be clinically
justified and measures should be taken to reduce the doses. The possible methods
include modification and optimization of scanning parameters and using
dose-reduction     techniques     (prospectively     electrocardiogram-gated      scan,
electrocardiogram-modulated tube current technology, etc).
                                                                   (Words Count: 477)

                                                   Signed:___________________

                                                           HUANG Bingsheng




                                           II
DECLARATION




I declare that this thesis represents my own work, except where due

acknowledgement is made, and that it has not been previously included in a thesis,

dissertation or report submitted to this University or to any other institution for a

degree, diploma or other qualifications.




                                                       Signed __________________

                                                                HUANG Bingsheng




                                           III
ACKNOWLEDGEMENTS


The research work for Mphil degree is very challenging and can only be based on
team work. I wish to take this opportunity to thank all the people who have helped
me for this.


First of all, I would like to thank my research supervisor, Dr. Khong Pek Lan. During
each step, including the literature review and project selection, doing the experiment,
analyzing the data, and writing the paper and thesis, Dr. Khong have been doing her
best to supervise me. I have been enjoying the studies under her supervision. I
believed that she is absolutely the best supervisor.


Secondly, I would like to thank Dr. Martin Wai-Ming Law, from Department of
Clinical Oncology, Queen Mary Hospital. He has given me valuable supervision and
advice during my research. He and his department also provided necessary
equipment for my experiment.


I have received great help from the staffs and students in the Department of
Diagnostic Radiology, the University of Hong Kong. I appreciate the help from
Wang Silun, Qiu Deqiang, Ms. Irene Leung and Miss Alice Lau. I am also very
grateful to the staff in PET-CT Unit in HKU, including Ken Liu, Chris, Thomas,
Winnie and Mr. Stephen Kwok for their kindly help with the experiments.


Last but not least, my sincere thanks should be given to my family, for their
continuous support and love for me in my life.




                                           IV
TABLE OF CONTENTS




TABLE OF CONTENTS


ABSTRACT................................................................................................................... I

DECLARATION..........................................................................................................III

ACKNOWLEDGEMENTS ........................................................................................ IV

TABLE OF CONTENTS ..............................................................................................V

LIST OF ABBREVIATIONS....................................................................................VIII

Chapter 1 Introduction ................................................................................................1

      1.1   Computed tomography (CT) technology and application ................................................1
           1.1.1 General principles of 64-Slice multi-detector CT (MDCT) ..................................1
           1.1.2 CT applications in cardiac scan.............................................................................4
      1.2 Positron emission tomography-computed tomography (PET-CT) technology and
      application .................................................................................................................................6
      1.3 Radiation dose from CT and PET-CT scan ....................................................................10
           1.3.1 Radiation dose from CT scan ..............................................................................11
           1.3.2 Radiation from PET-CT scan ..............................................................................13
      1.4 Cancer risk from radiation..............................................................................................14
      1.5 Research objectives ........................................................................................................15
           Figures:............................................................................................................................18
           Tables: .............................................................................................................................21
Chapter 2 Methodology ............................................................................................23

      2.1 Introduction .......................................................................................................................23
      2.2 CT dose measurement and estimation............................................................................24
           2.2.1 Direct-measurement using thermoluminescent dosimeters (TLD) on humanoid
           phantom...........................................................................................................................24
           2.2.2 CT dose estimation using ImPACT .....................................................................28
      2.3 PET dose estimation.......................................................................................................29
           2.3.1 PET dose estimation method in ICRP publication 80 .........................................29
           2.3.2 PET dose estimation method from MIRD report ................................................30
      2.4 Effective dose calculation ..............................................................................................31
      2.5 Cancer risk estimation....................................................................................................32
           2.5.1 Literature review .................................................................................................32
           2.5.2 Organ-specific risk estimation.............................................................................37
           2.5.3 Whole-body risk estimation ................................................................................37
      2.6 Comparison with baseline lifetime cancer incidence .....................................................38
           2.6.1 Baseline lifetime risk of cancer incidence in the HK population ........................38
           2.6.2 Baseline lifetime risk of cancer incidence calculation for the US population.....39
           2.6.3 Comparison between baseline lifetime risk and LAR of cancer incidence .........39
      2.7 Discussion of the limitations in methodology ................................................................40
      2.8 Summary and conclusion ...............................................................................................41
           Figures:............................................................................................................................42
           Tables: .............................................................................................................................49
Chapter 3 Cancer Risk Estimating Tables for the Hong Kong and United States


                                                                      V
TABLE OF CONTENTS


                 Population ................................................................................................59

    3.1  Introduction ....................................................................................................................59
    3.2  Methodology ..................................................................................................................60
        3.2.1 ERR model ..........................................................................................................60
        3.2.2 EAR model..........................................................................................................61
        3.2.3 Method for setting up the cancer risk estimating table for the HK population ...62
        3.2.4 Method for updating the cancer risk estimating table for the US population......63
    3.3 Results ............................................................................................................................63
        3.3.1 ERR for HK males and females ..........................................................................63
        3.3.2 EAR for HK males and females ..........................................................................64
        3.3.3 Cancer risk estimating table for the HK population ............................................64
        3.3.4 Updated cancer risk estimating table for the US population ...............................64
    3.4 Discussion ......................................................................................................................65
        3.4.1 Comparison between cancer risk estimating tables for the HK and the US
        population........................................................................................................................65
        3.4.2 Uncertainties of setting up the cancer risk estimating tables using BEIR VII
        report ...............................................................................................................................65
        3.4.3 Controversy of cancer risk estimation using BEIR VII report ............................66
    3.5 Summary and conclusion ...............................................................................................67
        Tables: .............................................................................................................................68
Chapter 4 Doses and Cancer Risks of pediatric cardiac CT Scan.............................81

    4.1     Introduction ....................................................................................................................81
    4.2     Literature review ............................................................................................................82
           4.2.1 Radiation exposure in cardiac pediatric CT scan ................................................82
           4.2.2 Cancer risk associated with cardiac pediatric CT scan........................................82
    4.3    Materials and methods ...................................................................................................83
           4.3.1 Organ-specific CT dose measurement.................................................................83
           4.3.2 Cancer risk estimation .........................................................................................84
    4.4     Results ............................................................................................................................84
           4.4.1 Radiation doses ...................................................................................................84
           4.4.2 LAR of Cancer Incidence....................................................................................85
    4.5     Discussion ......................................................................................................................86
    4.6     Summary and conclusion ...............................................................................................90
           Figures:............................................................................................................................91
           Tables: .............................................................................................................................93
Chapter 5 Doses and Cancer Risks of adult Coronary CT Angiography..................98

    5.1  Introduction ....................................................................................................................98
    5.2  Literature review ............................................................................................................99
    5.3  Methodology ................................................................................................................100
        5.3.1 Organ-specific CT dose estimation ...................................................................100
        5.3.2 Calculation of lifetime cancer incidence ...........................................................100
    5.4 Results ..........................................................................................................................101
        5.4.1 Radiation doses .................................................................................................101
        5.4.2 Cancer risks estimated for the HK and US population......................................101
        5.4.3 Comparison with baseline lifetime cancer risk..................................................102
    5.5 Discussion ....................................................................................................................102
    5.6 Summary and conclusion .............................................................................................104
        Figures:..........................................................................................................................105
        Tables: ...........................................................................................................................107
Chapter 6 Doses and Cancer Risks of Whole body PET-CT Scan .........................110

    6.1     Introduction ..................................................................................................................110
    6.2     Literature review ..........................................................................................................110

                                                                   VI
TABLE OF CONTENTS

      6.3 Materials and methods .................................................................................................111
          6.3.1 Organ-specific CT dose measurement and simulation ......................................112
          6.3.2 Organ-specific PET dose calculation ................................................................113
          6.3.3 Effective dose calculation .................................................................................113
          6.3.4 LAR of cancer incidence estimation for US and the HK population ................113
          6.3.5 LAR of cancer incidence compared to baseline ................................................114
      6.4 Results ..........................................................................................................................115
          6.4.1 Radiation doses .................................................................................................115
          6.4.2 Cancer risks .......................................................................................................116
      6.5 Discussion ....................................................................................................................117
      6.6 Conclusion....................................................................................................................120
          Figures:..........................................................................................................................121
          Tables: ...........................................................................................................................122
Chapter 7 Conclusion Summary and Future studies ...............................................131

      7.1     Conclusion summary....................................................................................................131
      7.2     Future studies ...............................................................................................................132
LIST OF FIGURES....................................................................................................140

LIST OF TABlES.......................................................................................................141

PUBLICATIONS AND PRESENTATIONS RELATED TO THE THESIS .............143




                                                                   VII
LIST OF ABBREVIATIONS


CT       computed tomography
MDCT     multi-detector computed tomography
EBCT     electron-beam computed tomography
MRI      magnetic resonance imaging
MRA      magnetic resonance angiography
CTA      computed tomography angiography
ECG      electrocardiogram
3D       three-dimensional
PET      positron emission tomography
PET-CT   positron emission tomography-computed tomography
FDG      fluorodeoxyglucose
SPECT    single photon emission computed tomography
CTDI     computed tomography dose index
ICRP     International Commission on Radiological Protection
US       United States
HK       Hong Kong
BEIR     Biological Effects of Ionizing Radiation Committee
TLD      thermoluminescent dosimeters
EAR      excess absolute risk
ERR      excess relative risk
MIRD     Medical Internal Radiation Dose
DDREF    dose and dose-rate effectiveness factor
LSS      life span study
LNT      linear no-threshold
LAR      lifetime attributable risk
ALARA    as low as reasonably achievable




                                      VIII
CHAPTER 1                                                                   BS HUANG




CHAPTER 1 INTRODUCTION


1.1 Computed tomography (CT) technology and application

1.1.1 General principles of 64-Slice multi-detector CT (MDCT)

After the principle of computed tomography (CT) technology was introduced in the

late 1960s, it has experienced great advancement, mainly including the improvement

of the spatial resolution (that is, how small an object size the CT scanner can

distinguish in an image) and temporal resolution (that is, how short a time the CT

scanner takes for imaging the object), and the reduction of radiation dose. However

the basic principles have been kept unchanged.



In CT scanners, X-Rays are produced in the X-ray tube in a standard way: by

accelerating electrons to a high energy (about 80-140 keV, depending on the tube

potential) and making them to collide with a metal target (normally tungsten). This

collision generates X-ray with an energy range from 0 keV to the energy of electrons

(80-140 keV). Metal filters are used to “remove” the low-energy X-ray photons

which contribute little to the production of medical images but cause unnecessary

radiation exposure to the human body. After being filtered, the X-ray passes through

a selected cross section of the patient from many angles, and is weakened from

absorption or scattering and deposited onto the patient. The X-ray beam which passes

through the human body is then detected by detectors and analyzed by electronic

circuit systems to obtain information such as energy and intensity, which represent

the information of the corresponding human body tissue and structure. The acquired

X-ray data are sent to computers, where the images of this human body section are
                                           1
CHAPTER 1                                                                   BS HUANG


reconstructed by applying reconstruction algorithms. In this way the images of all the

sections of the human body (or part of human body) are acquired and finally

displayed on the computer screen for diagnostic use.



Technological advancements in hardware and software such as in X-ray tubes, X-ray

detectors, electronic circuits and computer software have improved the performance

of CT scanners. For example, contemporary CT tubes have power ratings of up to

100 kW and anode heat capacity of 6 MJ (mega joule), yet retain an effective focal

spot area of less than 1 mm2. The dual source CT scanner released by Siemens has

higher combined power of 160 kW using two X-ray tubes (Siemens Healthcare,

2008). To date, CT scanner development has experienced five generations of X-ray

tubes and detectors.



The first generation CT scanner was invented in 1972 by Hounsfield. It had an X-ray

tube, linked to an X-ray detector located on the other side of the scanned object,

producing a narrow (pencil-width) beam of X-rays sweeping through the object

(Hounsfield, 1995). Compared to the first generation, the second generation designs

used 20 or more narrow beams and detectors reducing scan times to 20 s or less, but

the scanning process was similar to the first generation. For the third generation CT

scanners, the detector array and the X-ray tube were linked together in the gantry,

and rotated together around the patient when scanning. The X-ray beam was widened

into a fan-shaped beam. Many detector units were used (750 or more) to allow a

wide-angle measurement to be made (Goldman, 2007). The fourth generation CT

scanners incorporated a huge detector ring in the gantry with the X-ray tube alone

rotating around the patient (Goldman, 2007). However, as the detector arrays were


                                           2
CHAPTER 1                                                                    BS HUANG


more expensive than the ones in the third generation scanners, the fourth-generation

CT had only a very small market share, especially after the development of

multi-detector CT (MDCT) which is based on the third generation CT scanner. The

fifth generation CT scanners are known as the cine-CT with more than one X-ray

tube, and are only used for cardiac CT scan. It is however expensive and the

radiation exposure is high.



For better temporal resolution, the electron-beam CT (EBCT) was invented around

1984, with an electron beam (and consequently the focal spot) which is electronically

swept along all (or part) of the 360 circumference of the annular anode. EBCT has a

high temporal resolution and is able to perform a complete scan in as little as 10-20

ms, which made cardiac CT scan a reality due to the requirement of high temporal

resolution. However its application is limited to cardiac scans as the spatial

resolution of EBCT is relatively poor and the equipment is very costly. With

improvement of temporal resolution in MDCT, the future of EBCT is guarded

(Lembcke et al, 2006).



Most modern CT scanners in clinics are MDCT, based on a third-generation platform,

such as 4-slice, 16-slice and 64-slice (Goldman, 2008b), and even 256-slice scanners.

The main difference between single-slice and MDCT is the number of detector arrays

(in MDCT there are more than one detector arrays, while in single-slice CT there is

only one detector array). Hence this technology more effectively makes use of the

available X-ray beams as more data can be acquired in one rotation, and since the

number of rotations is reduced, X-ray tube usage is reduced. Current 64-slice MDCT

scanners, such as LightSpeed VCT (GE Healthcare, Milwaukee, WI), can cover a


                                           3
CHAPTER 1                                                                   BS HUANG


scan range of up to 40 mm (in one rotation) with slice thickness of 0.625 mm.



Both axial and helical scan modes, and even the cine mode, can be performed on the

current 64-slice CT scanners. In the axial scan mode, scan rotation is performed with

the table stationary and then the table is moved to the next location for another

rotation. In helical scan mode, the X-ray source rotates as the table moves the patient

through the scanner during an examination. Compared to the axial scan, the helical

scan markedly reduces the scanning time for the entire z-axis region of interest, and

in some cases within a single breath hold.



In MDCT, temporal resolution has been improved with short gantry rotation time and

low pitch factors. Modern CT scanners allow for gantry rotation of 0.3 seconds or

less. The dual-source CT system launched recently provides a high temporal

resolution of up to 60 ms (Flohr et al, 2006). Apart from temporal resolution, spatial

resolution has also improved dramatically. Nowadays, spatial resolution provided by

MDCT is about 0.35×0.35×0.5mm. Superior spatial resolution of CT scanners is an

important advantage compared to magnetic resonance imaging (MRI) (Matthew

Budoff et al, 2006), especially when small structures such as coronary arteries need

to be imaged. The spatial and temporal resolution of different imaging modalities for

angiography, including conventional invasive angiography, EBCT angiography,

MDCT angiography, and magnetic resonance angiography (MRA) are summarized in

Table 1.1 (Roberts et al, 2008).


1.1.2 CT applications in cardiac scan

Due to excellent spatial and temporal resolution, conventional angiography is still the


                                             4
CHAPTER 1                                                                     BS HUANG


gold standard for assessing the heart and coronary arteries. However, it is invasive

and may cause serious complications such as thrombo-embolism and artery

dissection. Therefore noninvasive methods of imaging, including CT angiography

(CTA) and MR angiography (MRA) can be advantageous. Applications of cardiac

CT and coronary artery CTA have been limited by problems such as heart motion,

respiratory motion and the small size of coronary arteries. Pre-requisites for CT

application in cardiac imaging are high spatial resolution and temporal resolution, as

the structure of the coronary artery is tiny and the heart beats at a high speed.

Technological advances with MDCT scanners have enhanced the spatial resolution

and temporal resolution.



However, usually it takes several seconds for the scanner to cover the whole heart,

depending on the different protocols and scanner types applied. This means that the

data used to reconstruct the cardiac CT images are acquired from several heart rate

cycles, and so the data may not be at the same heart beat phase. Applying the

electrocardiogram (ECG) –gated technology in cardiac CT scan, the data at the same

heart beat at diastole, can be selected and used for image reconstruction. Thus fewer

motion artifacts are produced in the cardiac CT images (Desjardins et al, 2004) and

thus, the application of CT in angiography has increased. There are two kinds of

ECG –gated technologies, namely, retrospective ECG gating and prospective ECG

triggering. For retrospective ECG gating, data are acquired during the whole heart

cycle, and only part of that (data at the diastole heart beat phase) is used for image

reconstruction. In order to obtain enough raw data, data oversampling is used with a

low pitch factor which depends on the patient’s heart rate. For prospective ECG

triggering, initially the mean duration of a heart beat cycle is averaged from multiple


                                            5
CHAPTER 1                                                                    BS HUANG


heart cycles. The trigger, a predefined time point in each other subsequent cardiac

cycle, is used to “trigger” a sequential axial scan during the diastole heart beat phase.

The acquisition time for one axial position is about 250–500 ms. All the data

acquired in prospective ECG triggering scan are used for data reconstruction.



The clinical applications of cardiac CT in adults include calcium scoring, coronary

angiography, evaluating atherosclerotic coronary plaques, coronary stents and

anomalies, etc (Prat-Gonzalez et al, 2008). Cardiac CT scan has been applied in

children for detection of pathology such as congenital heart diseases, great vessel

anomalies, intracardiac shunt lesions, post-operative anatomy and less frequently, for

the evaluation of coronary artery anomalies and coronary artery aneurysms in

Kawasaki’s disease(Gilkeson et al, 2003; Westra et al, 1999). Figure 1.1 shows the

three-dimensional (3D) coronary CTA image captured with the GE healthcare

64-Slice LightSpeed VCT scanner.


1.2 Positron emission tomography-computed tomography (PET-CT)
technology and application

In positron emission tomography (PET) scan, biologically active molecules such as

fluorodeoxyglucose (FDG) labeled with a radioactive tracer isotope such as 18F, that

is 18F-FDG, are injected into the blood circulation of the patient. As the radioisotope

undergoes positron emission decay (also known as positive beta decay) it emits a

positron, the antimatter counterpart of electron. The half life of this decay is 110

minutes, which is long enough such that fluorine-18 labeled radiotracers can be

manufactured commercially at an offsite location, and short enough such that the

radioisotopes decay soon and do not harm the patients. After traveling for a short


                                            6
CHAPTER 1                                                                     BS HUANG


distance (less than a few millimeters) the positron encounters and annihilates with an

electron, producing a pair of annihilation photons (γ-ray) moving in opposite

directions with a kinetic energy of 0.511 MeV. The electronics system detects and

records pairs of γ-ray signals, but discards the single photon signal. From the

information of γ-ray, the distribution of labeled biological molecules in human body

is acquired, and this reflects functional information of human tissue. However, one or

both of these two photons may be lost due to their absorption in the patient’s body or

their scattering out of the detector field of view. This effect is called attenuation, and

is greater with PET imaging compared to traditional nuclear medicine single photon

emission computed tomography (SPECT) imaging, because two photons must escape

the patient simultaneously to be detected as a true event. Attenuation effect increases

image noise, image artifacts, and image distortion in PET images. Therefore,

attenuation correction of PET scan data is necessary. On the traditional stand-alone

PET scanner, attenuation correction is done by using a gamma ray (positron emitting)

source, such as 68Ge, and the detector arrays of PET scanner. From the data acquired

with the detectors, the attenuation effect can be quantified and used for PET scan

data correction (Blodgett et al, 2007).



PET examinations using FDG (FDG –PET) have been widely used for the diagnosis

and staging of cancer, and for monitoring the treatment of cancer. FDG is a glucose

analog that is phosphorylated by hexokinase (whose mitochondrial form is greatly

increased in malignant tumours) in glucose-using cells. In FDG, one oxygen atom is

replaced by 18F and as the oxygen atom is required for the next step in glucose

metabolism, no further reactions occur in FDG. Furthermore, most tissues cannot

remove the phosphate added by hexokinase. This means that 18F -FDG is trapped in


                                            7
CHAPTER 1                                                                      BS HUANG


any cell which takes it up until it decays. This results in intense radioactivity in

tissues with high glucose uptake, such as brain, liver, and most cancers. This

provides functional information of the human body. On the other hand, CT is still a

primary tool of anatomical imaging and superior in spatial resolution to the PET scan.

Therefore, the fusion of anatomical and functional images is advantageous, and this

was first realized by using software methods of co-registration. However, the

alignment procedure for functional images and anatomical images is tedious and

lacks accuracy. Hence, it is valuable to acquire both functional information and

anatomical information at the same examination, with accurately aligned images.

PET-CT scanner, the merger of PET and CT scanners satisfies this need with fusion

of anatomical images from CT scan and functional images from PET scan (Beyer et

al, 2000). Example of PET-CT scan is shown in Figure 1.2.



In the PET-CT scanner, all the components of the PET scanner and CT scanner are

assembled within one single rotational gantry with a distance of about 60 cm

between the two scanners’ centers. The combined PET-CT scanner can run in CT

mode or PET mode alone, or a combined mode in which the CT images are used for

attenuation correction of PET data and fusion.



A typical PET-CT scan starts with an injection of FDG followed by a 60 minute

uptake period. The patient is firstly scanned with a scout scan to determine the axial

range of the helical CT scan, and then a helical CT scan is performed. Once CT scan

is completed, the patient is moved to the first bed position for PET scan. A

whole-body emission PET scan of the same transverse coverage and the same

contiguous segments as CT scan is obtained with several minutes’ acquisition per bed


                                            8
CHAPTER 1                                                                 BS HUANG


position. For example, in the PET-CT unit of the University of Hong Kong, it

normally takes 2.5 minutes for scanning one bed position and 17.5 minutes for whole

body with 7 bed positions. The data of CT images is sent to the PET console for PET

data scatter and attenuation correction before PET image reconstruction (Beyer et al,

2000). The reconstructed PET image is co-registered with the CT image by simply

accounting for the shift of two scanning centers, and the fused images are displayed

on the computer screen.



The technology of CT-based attenuation correction for the PET images not only

enhances the quality of PET images compared to conventional PET scans which uses

a γ-ray (positron emitting) source for attenuation correction, but also reduces scan

time(Beyer et al, 2000), although the latter has the advantage of increased accuracy

due to emission of the same gamma ray energy and therefore the same attenuation

coefficient map as the emission scan (Kinahan et al, 2003). Due to the advantages

compared to CT and PET alone, the application of PET-CT scanners have been

expanding in oncological management such as diagnosis and staging of malignant

disease, image-guided therapy planning, treatment monitoring, and also in the

investigation of fever of unknown origin. Since the advent of PET-CT, many studies

have been done to evaluate its diagnostic performance. In the study of Bar-Shalom

and colleagues, it was found that PET-CT provided additional information over the

PET and CT alone in 49% of patients and        30% of the sites, and PET-CT had a

substantial impact on the management in 14% of the patients (Bar-Shalom et al,

2003). Ambiguous lesions were found in only 3.4% of the lesions by PET-CT

compared to 15.3% in PET alone group (Pelosi et al, 2004).




                                          9
CHAPTER 1                                                                     BS HUANG


To date, the expanding clinical applications of PET-CT have led to an increasing

demand for PET-CT scans and more combined PET-CT scanners installed in

hospitals and clinics worldwide.


1.3 Radiation dose from CT and PET-CT scan

The energy deposited by radiation causes “absorbed dose” in the human body,

defined as radiation energy deposited in the medium of unit mass, as in Equation

(1-1)

                                D = dE / dm                                     (1-1)

Where dE is the energy deposited into a small medium and has the unit of joule (J);

dm is the mass of this medium with the unit of kilogram (kg); D is absorbed dose and

has the unit of J/kg (1 joule per kilogram), which is given the unit Gy. The energy

deposited in various organs or tissues is generally different because the thickness and

attenuation ability are different, and so are the absorbed doses. Actually the various

kinds of ionizing radiation such as X-ray, γ-ray, or α particle have different effects on

the same human tissue due to the different abilities of ionizing, and therefore “the

equivalent dose” is defined as in Equation (1-2)

                              H T = ∑ WR ⋅ DT , R                           (1-2)
                                      R


where WR is the radiation weighting factor without any unit (the value depends on the

type of radiation, such as photon, electron, neutron or proton, etc); DT,R is the

absorbed dose in tissue T or organ T; HT has the same unit as DT,R (J/kg), which is

however given a special unit sievert (Sv). Usually this unit is too large for radiation

protection dose range, so milli sievert (mSv, 0.001 Sv) is used. For photons (such as

X-ray and γ-ray) WR is 1 and so in CT or PET scans, equivalent dose is equal to

absorbed dose. The definition of equivalent dose is for the comparison of doses from

                                           10
CHAPTER 1                                                                     BS HUANG


different radiation types.


1.3.1 Radiation dose from CT scan

As described above, in CT X-rays pass through a cross-section of the patient and are

weakened because some energy is deposited into the patient. The X-ray used in CT

scan often has an energy range of 0 - 80 keV, 0 - 100 keV or 0 -120 keV depending

on the tube potential used, and most of the X-ray energy is around 1/3 of the tube

potential. The radiation dose characteristics of a CT scanner are often described in

terms of computed tomography dose index (CTDI) (Shope et al, 1981), which is

defined as the dose measured from 14 contiguous sections in a single axial scan and

normalized to beam width, as shown in Equation (1-3)

                                                7T
                             CTDI = (1/ nT) ∫ Dsin gle (z)dz   (1-3)
                                                −7T


where n is the number of slices, T is the section thickness, and Dsingle(z) is the dose at

point z on the z axis. The product of n×T is used to reflect the total nominal width of

the X-ray beam during acquisition. However exposure is often measured using a

pencil ionization chamber, whose fixed length of 100 mm is not equal to the length

of 14 CT slice sections. Therefore CTDI100 is defined (European Study Group of

Radiologists and Physicists, 2008), as in Equation (1-4)

                                         +5cm
                    CTDI100 = (1/ nT) ∫ Dsin gle (z)dz           (1-4)
                                         −5cm


Where n, T, and z has the same definition as in Equation (1-3). Usually two

polymethyl methacrylate phantoms (one with diameter of 32 cm for trunk; the other

with diameter of 16 cm for head) are used for CTDI measurements. For different

position in the scan range, the CTDI (or CTDI100) is different. To avoid this shortage,


                                                      11
CHAPTER 1                                                                     BS HUANG


CTDIw is used to provide a weighted average of the CTDI100 in the center and

CTDI100 in periphery (European Study Group of Radiologists and Physicists, 2008),

as shown in Equation (1-5)

                            1            2
                                central + CTDI
                     CTDIw = CTDI             peripheral          (1-5)
                            3            3

For helical CT scan, CTDIvol is defined, as in Equation (1-6)

                    C TDI vol = CTDI w / pitch                  (1-6)

This CTDIvol descriptor takes into account the parameters that are related to a

specific imaging protocol, the helical pitch or axial scan spacing. CT is recognized as

producing much more radiation dose than other diagnostic modalities such as

mammography (Aroua et al, 2002; Brugmans et al, 2002), and cardiac CT

examinations especially coronary CTA (Einstein et al, 2008) generally produce more

dose than other CT scan types, such as chest CT (Huda, 2007) and abdominal CT

(Ware et al, 1999). The 64-slice MDCT even produces higher radiation dose than

16-slice CT for the same coronary CTA (Dewey et al, 2007). The doses from cardiac

CT scan may be reduced by ECG-modulated tube current. This technique provides

full tube current to the patient only in the diastolic period of the cardiac cycle, during

which it is most likely to produce the best image quality. The tube current is

modulated to a lower mA setting during systole to decrease the dose to the patient

(Einstein et al, 2007; Hermann et al, 2008; Kalender et al, 1999). This technique is

shown in Figure 1.3.



However the radiation doses of CT scans to human organs are still high despite the

techniques being applied. Table 1.2 summarizes examples of the typical organ doses

from various radiologic studies.



                                           12
CHAPTER 1                                                                BS HUANG


1.3.2 Radiation from PET-CT scan

Radiation dose from PET and CT are different in some aspects. Firstly, the ionizing

radiation of PET scans is from γ-rays with a higher energy of 511 keV, while

radiation of CT scanners it is from X-rays with a lower energy of at most 140 keV.

Secondly, X-ray is generated in the CT scanner only when the scanning is on-going

and of a limited duration, while γ-ray in PET is associated with the patient from the

time he/she is injected with the radiopharmaceuticals, until the radiopharmaceuticals

decay to the background level. Thirdly, the human body is exposed to the X-ray from

an external source, while the radiopharmaceutical γ-ray is generated internally.

Combined PET-CT examinations result in an increased patient radiation exposure

compared to a conventional CT or PET examinations, as dose of PET-CT scan is

composed of combined dose of PET scan and CT scan.




                                         13
CHAPTER 1                                                                     BS HUANG



1.4 Cancer risk from radiation

Not long after the discovery of X-rays in 1895 and of natural radioactivity in 1896,

clinical evidence, mainly from effects on the skin, indicated that ionizing radiation is

harmful to human tissues. Later it was realized that not only is ionizing radiation

damaging to most tissues but exposure of the germinal tissue in plants and animals

was found to result in effects in the descendants as well. After over more than a

century of exploring the uses and adverse effects of ionizing radiation, it became

evident that human beings must study the biological effects of ionizing radiation in

order to protect themselves and other species from its harmful effects while at the

same time maximizing the benefits of its use.



Biological effects on human beings can be categorized into deterministic effect and

stochastic effect according to International Commission on Radiological Protection

(ICRP) publication 60 (ICRP, 1991). According to experiments and theoretical

studies, the severity of deterministic effects on human beings is absent below a

certain level (the “threshold”) and increases with increasing doses above the

threshold. For stochastic effect, severity is independent of the absorbed dose. There

is no dose threshold for the stochastic effect, but the chance of having the effects is

proportional to the absorbed dose. Examples of deterministic effect include cataract

and erythema, while stochastic effects include radiation induced cancer risk and

genetic effect. The threshold of deterministic effect is often at a level of 1 Gy, which

is much higher than the radiation dose from general diagnostic radiology (Table 1.2),

which is in the range of 1 mSv to around 100 mSv. Hence in our study, we mainly

focus on the stochastic effect, especially cancer risk induced by the radiation.


                                           14
CHAPTER 1                                                                  BS HUANG




The same radiation exposure to various organs and tissues has different probabilities

for the occurrence of stochastic effects. This different sensitivity to stochastic

damage is reflected by the tissue weighting factors according to ICRP publication 60

and ICRP publication 103 (ICRP, 1991; ICRP, 2007). For organs which are more

likely to have stochastic damage, the factors are correspondingly higher. To express

the organ-specific doses in terms of an equivalent uniform dose to the whole body,

the effective dose is calculated by summing up the organ doses weighted by

corresponding tissue weighting factors (details are introduced in Section 2.4).


The same ionizing radiation generally results in imparting a higher radiation dose to

children than to adults. This is due to the smaller weight of children even though the

radiation deposits much more energy into adults, according to Equation (1-1). For

example, from an abdominal CT scan, only 72 mJ (millijoule) are deposited in

children instead of 235 mJ in adults due to the smaller volume of tissues exposed.

However, the effective dose is 6.1 mSv in children and only 3.9 mSv in adults (Ware

et al, 1999). Similarly, for a head CT scan, effective dose is 3.7 mSv in children and

only 1.0 mSv in adults (Huda, 2002). Moreover, for the same effective dose, children

will suffer higher biological effects and lifetime risks for than adults (ICRP, 1991).

This is because the organs and tissues of children are more sensitive to the effects of

radiation than adults, and children have a longer life expectancy than adults.

Therefore, CT scanning parameters that control radiation dose should be tailored for

children.


1.5 Research objectives

In Hong Kong, cancer has been the leading cause of death from 1990 to 2005

                                          15
CHAPTER 1                                                                      BS HUANG


(Hospital Authority of Hong Kong, 2008). It is foreseen that the use of PET-CT scan

in patients with cancer will exponentially increase in the future due to its increasing

role in the management of cancer patients.



According to the Hospital Authority of Hong Kong, heart diseases including

hypertensive heart disease, is the second leading cause of death from 1990 to 2005

(Hospital Authority of Hong Kong, 2008). Hence cardiac CT has become

increasingly important for diagnosing heart disease, especially with the advent of

64-slice CT scanner (Wintersperger et al, 2005).



These imaging modalities impart ionizing radiation to the individual patients and also

contribute to the increased radiation dose to the general public. Hence, it is necessary

to study radiation dose and the associated cancer risks, for education and providing

justifications to the individual patient and for the public. In our study, we firstly

aimed to measure and calculate the radiation dose to organs using an

anthropomorphic phantom, resulting from (1) 64-slice cardiac CT scan in pediatric

patients , (2) 64-slice coronary CTA in adult patients, (3) whole body PET-CT scan

in adult patients. We also aimed to estimate the cancer risk caused by these radiation

doses, for both the United States (US) and Hong Kong (HK) populations. It was

believed that the comparison of the results between the two populations would be

relevant and interesting, and the study will be of impact to a larger community.

Estimations for the HK population could also be extended to the South China

population due to similar demographic factors.




                                            16
CHAPTER 1        BS HUANG




            17
CHAPTER 1                                                       BS HUANG



Figures:




  Figure 1.1:   Three-dimension coronary CT angiography image captured
                with the GE Healthcare 64-slice LightSpeed VCT scanner.




                                  18
CHAPTER 1                                                        BS HUANG




          (CT)                       (PET)                (PET-CT)

Figure 1.2:   CT, PET and PET-CT images (Taken in the PET-CT Unit of The
              University of Hong Kong.) showing a hypermetabolic left lower
              lobe lung mass and multiple hypermetabolic foci in the axial
              skeleton and pelvis.




                                      19
CHAPTER 1                                                          BS HUANG




Figure 1.3:   Electrocardiogram (ECG)-modulated tube current technique. In
              this example, start phase is 70%, end phase is 80%, namely, the
              tube current is maximum from 70% to 80% of the heart beat cycle,
              and lower for the rest. This image is from GE LightSpeed VCT
              Technical Reference Manual.




                                      20
CHAPTER 1                                                            BS HUANG



Tables:


                                    spatial resolution   temporal resolution
 study type
                                          (mm)                 (ms)
 conventional angiography                   0.2                5–20
 electron beam CT                          >0.6               33–100
 16-Slice CT                                0.5                 200
 64-Slice CT                                0.4                 165
 dual-source 2×64-slice CT                  0.4                 83
 magnetic resonance angiography             0.7                 20

Table 1.1:    Comparison of temporal and spatial resolution of typical CT
              scanners and other study types used for angiography.




                                      21
CHAPTER 1                                                           BS HUANG



                                                         relevant organ dose
study type                             relevant organ
                                                            (mGy or mSv)
dental radiography                          brain               0.005
posterior–anterior chest radiography        lung                 0.01
lateral chest radiography                   lung                 0.15
screening mammography                      breast                  3
adult abdominal CT                        stomach                 10
neonatal abdominal CT                     stomach                 20
CT angiography                         breast, stomach          50~70

Table 1.2:    Typical organ doses from various radiologic studies. All data are
              from the work of Brenner and Hall (Brenner et al, 2007), except
              CT angiography which is from the work of Einstein et al (Einstein
              et al, 2007).




                                       22
CHAPTER 2                                                                    BS HUANG


CHAPTER 2 METHODOLOGY


2.1 Introduction

In this chapter, the main methodologies applied in our studies are introduced and

discussed, including the methodology for measuring or simulating organ doses from

CT scan, for estimating the doses from PET scan, for calculating the effective dose,

and for estimating lifetime risks of cancer incidence for the United States (US) and

Hong Kong (HK) populations. The methods introduced in reports of authoritative

radiation protection institutions for estimating biological effects of radiation were

reviewed and compared, and the principles introduced by Biological Effects of

Ionizing Radiation (BEIR) Committee and the National Research Council of USA in

its BEIR VII report were applied to estimate the cancer risk associated with the doses

from CT or PET-CT scan in our studies. Also described in this chapter is how we

compared the calculated lifetime risk of cancer incidence with the baseline lifetime

cancer incidence. Uncertainties and limitations of the methods for estimating

radiation dose and cancer risk are also discussed.




                                           23
CHAPTER 2                                                                     BS HUANG



2.2 CT dose measurement and estimation

2.2.1 Direct-measurement using thermoluminescent dosimeters (TLD)
on humanoid phantom

2.2.1.1 Adult dose


The radiation dose from CT scan for adults was measured on a Rando phantom

(Alderson Research Laboratories Inc., Long Island City, New York) equipped with

LiF thermoluminescent dosimeters (TLD) with dimensions of 3.2 mm×3.2 mm×

0.9 mm (type TLD-100, Harshaw Chemical Company, Solon, Ohio, USA). The

phantom used in the experiment was a female phantom representing a 163 cm (5’4”)

tall and 54 kg (118 lb) female figure, which was converted to a male phantom by

removing the breast attachments (Figure 2.1). The humanoid Rando phantom is made

of 34 slices with a slice thickness of 2.5cm. The slices are constructed with

radiologically equivalent tissues including soft tissue, lungs, skeletons and breast

cups. Holes are drilled through the slices, which allow standard plugs to be inserted.

The plugs can be used to hold a variety of dosimeters, such as TLD.



TLD is widely used for radiation measurement. When ionizing radiation interacts

with TLD, all or part of the initial energy of the radiation is deposited in TLD. Some

of the atoms in TLD that absorb that energy become ionized, producing free

electrons and “holes” (created when the electron is ionized away from the initial

position leaving a   hole) trapped in the crystal structure (lattice) of the TLD.

Heating the TLD causes the lattice to vibrate, releasing the trapped electrons (or

holes) in the process if the temperature is high enough. Released electrons (or holes)


                                           24
CHAPTER 2                                                                   BS HUANG


return to the original ground state and recombine with holes (or electrons), releasing

the captured energy as light (photons). These photons can be counted using the

detector in the TLD reader. The number of photons counted is proportional to the

amount of energy deposited in the crystal, and therefore, readings from the TLD

reader reflect the amount of deposited energy. The electrons and holes in original

ground state are excited and trapped again if TLD is irradiated, hence the TLDs can

be re-used. Samples of TLD-100 which are used in our study for adult dose

measurement are shown in Figure 2.2.



To remove the inferior TLD chips, the TLD-100 chips which were newly bought

were irradiated by a linear accelerator in the Department of Oncology (type 4-EX,

Virian, Palo Alto, California). Mean and standard deviation of the TLD-100 readings

were calculated. The chips with readings which were outside the range of ±5% were

discarded (TLD selection). Hence the inherent precision of the readings of our

TLD-100 used for dose measurement is ±5%. In addition, the directionality error

associated with the edge and surface of chips has been calculated to be ±2% (Wagner

et al, 2000). Therefore the final TLD-100 readings have an error of ±7%.



Before being used for dose measurement, the TLD-100 chips were calibrated using a

CT scanner (Discovery VCT system, GE Healthcare Systems, Milwaukee, WI, USA)

and an ionization chamber (10X5-3CT, Radical Corporation, Monrovia, CA, USA).

The chamber can be used for measuring the air kerma (results are in milliRoentgen,

mR). The chips were divided into subsets, which were placed into black plastic bags

and attached to the ionization chamber. The chamber attached with chips was

scanned by the CT scanner with general scanning parameters. After correction for


                                          25
CHAPTER 2                                                                   BS HUANG


noise of TLD-100 readings, the relationship between the air kerma from the

ionization chamber and the readings from TLD-100 chips was found to be linear and

a calibration factor was derived from this relationship, as shown in Table 2.1.



The calibration results are given as

                             air kerma (mR) = a*charge (nC)                   (2-1)

where a is 13.3 (±0.2) mR/nC for 120 kV, or 13.0 (±0.1) mR/nC for 140 kV.



Before being inserted into the phantom, the TLD-100 chips were pre-heated using an

annealing oven (TLD annealing oven, PTW-Freiburg, Freiburg, Germany) (Figure

2.5) to release the chips to the original state, which makes the TLD chips reusable.

The organs selected to measure the doses were the ones for which tissue weighting

factors are recommended by ICRP publication 103 (ICRP, 2007). At least two chips

were used for a specific organ, and the readings were averaged to calculate the organ

dose. Using more than 1 TLD chips may lower the uncertainty of the measured

values and prevent the possible damage to TLD chips. However, the positions of the

chips may cause additional error to the organ dose results as radiation exposure

varies with locations. Twenty-four hours after being irradiated by the CT scanner, the

TLD-100 chips were read by a TLD reader (Harshaw, model QS5500, OH, USA)

(Figure 2.6). This time period (24 hours) was applied for TLD fading correction

(Maria Ranogajec-Komor, 2003). The results from the TLD reader were converted

into dose values (in mR) by multiplying the readings with the factors acquired from

the TLD calibration. Dose measurement results from the TLD-100 system were

expressed in mSv based upon transfer factors from mR to mSv ( in air it is 0.0087

mSv per mR) (A.R.Lakshmanan, 1991).


                                          26
CHAPTER 2                                                               BS HUANG


2.2.1.2 Pediatric dose

The radiation dose of pediatric cardiac CT scan was measured with a standard

5-year-old pediatric phantom (CIRS, model 705-C, Norfolk, VA, USA) equipped

with LiF 3.2 mm×3.2 mm×0.6 mm TLD-100H (type TLD-100H, Harshaw Chemical

Company, Solon, Ohio, USA). The pediatric phantom is composed of 26 slices with

a slice thickness of 2.5 cm, each of which contains holes for TLD placement (Figure

2.3). In this study   TLD-100H but not TLD-100 chips were used as TLD-100H are

more accurate for measurement of low radiation dose exposure to some organs which

were not directly exposed to the X-ray in cardiac CT scan (such as brain, bladder,

gonads, etc). The same procedure of selection from the newly bought chips was

carried out for TLD-100H as for TLD-100, and the variation of TLD-100H chips

reading was also 5%. The final TLD-100H readings have also an error of ±7%

considering the directionality error.



The TLD-100H used in the study were also calibrated using a 100 kVp X-ray beam

from the superficial X-ray machine (Philips, RT100, Germany). The other equipment

and methods were the same as calibration of TLD-100. After correction for noise the

calibration factor was calculated to be 0.679 mR/nC. The method of processing

TLD1-00H is also the same, except the different calibration factor.



To summarize, the complete process of dose measurement using TLD is illustrated in

Figure 2.4.




                                         27
CHAPTER 2                                                                   BS HUANG


2.2.2 CT dose estimation using ImPACT

In addition, doses of CT scan was simulated using a CT patient dosimetry calculator

— ImPACT (Medical Devices Agency, 2006) (Figure 2.7), which is a widely

recognized tool for calculating patient organ and effective doses from CT

examinations. The spreadsheet makes use of the National Radiological Protection

Board (NRPB) Monte Carlo dose data published in report SR250 (Jones DG, 1991),

which provides normalized organ dose data for irradiation of Medical Internal

Radiation Dose (MIRD) phantom (Snyder et al, 1969) by a range of old CT scanners

produced before 1993. For the new scanners produced after 1993, the normalized

organ doses data have been updated in the spreadsheet by matching old CT scanner

data in SR250 with new CT scanner dose survey data (CTDI, etc), but direct Monte

Carlo simulations were not made.



For our studies, the type of 64-slice MDCT scanner and CT protocol parameters were

entered into the spreadsheet. As ImPACT requires inputting fixed values, for

protocols using automA technology (for example, the CT part in PET-CT scan), the

upper and lower limits of current values were input, because the exact tube current at

each specific body region is unknown. Input of these values (upper or lower limit)

would give us the results for the dose range. The spreadsheet makes use of these

input data and the Monte Carlo simulation results from National Radiological

Protection Board (Jones DG, 1991) to determine the organ doses. Effective dose was

computed in ImPACT using the tissue weighting factors recommended in ICRP

publication 60. We then recalculated effective dose according to the weighting

factors from ICRP publication 103 which is more updated, and to maintain

consistency with other effective dose calculations in our study (refer to section 2.4

                                          28
CHAPTER 2                                                                    BS HUANG


Effective dose calculation).


2.3 PET dose estimation

                                                                               18
There are mainly two models used for estimating the radiation dose from             F-FDG

PET scan described in the literature: including the model recommended by ICRP

publication 80 and another by MIRD committee.


2.3.1 PET dose estimation method in ICRP publication 80

ICRP presents dose coefficients for calculating doses from 18F-FDG PET scan, which

is acquired by multiplying the FDG activity with the coefficients recommended in its

publication 80. These coefficients are for a variety of organs and tissues of the human

body, as described in Table 2.2. The residence kinetic data, which describes the

change of radiotracer activity in the organ of interest with time, forms the basis of

PET dose calculation. The coefficients in ICRP publication 80 are based on the FDG

kinetic data from the study of Jones et al (Jones et al, 1982), which assumed that 4%

and 6% of the administered activity were taken up by the myocardium and brain

respectively and all other activity was distributed uniformly in the body.



Using the dose coefficients by ICRP publication 80, the organ-specific dose to

patients from the PET scan can be calculated by multiplying the factors with the

injected 18F-FDG activity

                                    D TPET = A * ΓTFDG

         PET
where DT       stands for the organ dose from PET scan, A stands for the injected

activity, and ΓTFDG is the recommended coefficient, which is an average value for

men and women.

                                           29
CHAPTER 2                                                                  BS HUANG


2.3.2 PET dose estimation method from MIRD report

The MIRD Committee also recommended organ-specific dose coefficients to
                                18
estimate the organ doses from        F-FDG PET examinations in its publication MIRD

Dose Estimate Report No. 19 (Hays et al, 2002) In the MIRD report, the kinetic data

are based on four models, including the one by Hays and Segall (Hays et al, 1999),

the one for 18F-FDG in healthy Japanese subjects by Mejia et al (Mejia et al, 1991),

the one for 18F-FDG in bladder in the study by Jones et al (Jones et al, 1982), and the

one for 18F-FDG in brain by Niven et al (Niven et al, 2001). Weighted means of these
                                                                              18
four were used for estimating the organ absorbed doses from                        F-FDG

administration.



The differences between the coefficients provided in ICRP publication 80 and MIRD

committee dose report are mainly due to the difference between the residence kinetic

data used in the two publications. For example, the brain dose in MIRD report is

0.046 mGy per MBq, which is higher than 0.011 mGy per MBq in ICRP publication

80. The difference in brain dose coefficients between the two reports is due to the

different brain residence times (MIRD report uses a brain residence time of 0.23 h,

which is higher than the ICRP value of 0.15 h).



It is hard to judge which one of the two methods is better, since both of them are well

accepted and used in studies about dose from FDG PET scans. However, we decided

to calculate the “effective dose” (as introduced below) using the tissue weighting

factors recommended in ICRP publication 103, and these factors are applied for the

same organs for which dose coefficients are provided in ICRP publication 80. In the

MIRD report, the “total-body” dose, which is acquired by dividing total energy

                                            30
CHAPTER 2                                                                   BS HUANG


deposited in the body with total body mass, is totally different from the term

“effective dose” defined in ICRP publications and cannot be directly compared with

effective dose. Hence, we applied the method introduced in ICRP publication 80 to

estimate the organ doses and effective dose from FDG PET examinations.


2.4 Effective dose calculation

According to the knowledge of radiation protection, biological effects are not solely

dependent on the absolute radiation dose, but also on the organs irradiated. To

describe the impact of organ-specific dose on health, the tissue weighting factors

which were defined by ICRP firstly in its publication 60 and redefined in publication

103 were used. The term of “effective dose” was also introduced in ICRP publication

60 to calculate a uniform whole body dose that in theory gives the same risk as the

nonuniform dose pattern that actually occurs. This provides the possibility of

comparing risks from different nonuniform radiation which may be imparted to

different human organs. The effective dose is the sum of the weighted equivalent

doses in all the tissues and organs of the body, and is given by

                                     E = ∑WT ∗HT
                                         T



where E is the effective dose, H T stands for organ-specific dose, WT is the tissue

weighting factor for a specific organ or tissue T as listed in ICRP publication 60 or

103. The tissue weighting factors from ICRP publication 60 and 103 are shown in

Table 2.4.



What should be noted is that the tissue weighting factors have been developed from a

reference population of equal numbers of both sexes and a wide range of ages.

Therefore the definition of tissue weighting factors and effective dose can be applied
                                             31
CHAPTER 2                                                                   BS HUANG


to workers exposed to radiation or radioactive substances at work, and the world

population across different ages and both genders.



In our research, the tissue weighting factors from ICRP publication 103, not ICRP

publication 60, were applied for calculating the effective dose. This is because these

weighting factors are updated values compared to those in ICRP 60. The new tissue

weighting factors recommended in ICRP publication 103 are based on updated

information from epidemiological studies on cancer induction and risk assessment

for hereditary effects. The following changes were made in ICRP 103:

1. Risks from hereditary effects and cancer induction from gonadal irradiation were

considered and a WT of 0.08 was given to replace 0.12.

2. Cancer risks of brain were considered to be higher than expected in ICRP

publication 60 and a WT of 0.01 were given to brain instead of 0.005.

3. A greater WT of 0.12 was given to breast due to the higher risk in breast cancer in

females instead of 0.05.


2.5 Cancer risk estimation

2.5.1 Literature review

Radiation from the Atom-bomb explosion in Japan during World War II has caused

terrible diseases and deaths. However, it has provided valuable information for

researchers who study radiation risk. Many methods have been developed to estimate

the risk of radiation exposure to human beings after realizing the possible harms that

may occur. In this chapter, we review these radiation-induced cancer risk estimating

methods, including those from National Research Council BEIR Committee,

International Commission on Radiological Protection (ICRP), National Council on

                                          32
CHAPTER 2                                                                  BS HUANG


Radiation Protection & Measurements, US Environmental Protection Agency, and

United Nations Scientific Committee on the Effects of Atomic Radiation.



Two cancer risk estimating models, excess absolute risk (EAR) model and excess

relative risk (ERR) model, are discussed in these reports. EAR is the rate of cancer in

an exposed population minus the rate of cancer in an unexposed population (baseline

cancer rate), while ERR is the rate of cancer in an exposed population divided by the

disease in an unexposed population minus 1. The EAR model assumes that excess

risks caused by radiation remain constant between different populations, and do not

depend on baseline risks whilst the ERR model assumes that the excess risks are

proportional to baseline risks of different populations.


2.5.1.1 Cancer risk estimating methods from publications other than BEIR VII

BEIR IV report from NRC established a model for lung cancer as a function of time

since exposure and attained age from data based on four cohorts of underground

miners using ERR model (NRC BEIR Committee, 1988). It was the first major

radiation risk assessment based on ERR model. BEIR V report (NRC BEIR

Committee, 1990) also developed ERR models for estimating cancer risk from

low-dose radiation as a function of gender, exposed age, radiation dose, and time

since exposure. The ERR models for mortality risk of breast cancer, respiratory

cancer, digestive cancer and all other solid cancers were developed from the data of

Atom-bomb survivor mortality data in the period 1950-1985 and Canadian

fluoroscopy patients (for breast cancer only). ERR models were also set up for breast

and thyroid cancer incidence. The ERR models were expressed as a linear function of

dose after modification by factors, duration of exposure, sex, exposed age, and time



                                           33
CHAPTER 2                                                                    BS HUANG


since exposure. For low dose radiation, a reduced risk was recommended by BEIR V

report by dividing the estimates with a factor, namely, dose and dose-rate

effectiveness factor (DDREF) of between 2 and 10. The US population data in 1980

were applied to calculate lifetime cancer risk estimate of three different scenarios: a

single exposure of 0.1 Sv, continuous lifetime exposure to 1 mSv per year, and

continuous exposure to 1 mSv from age 18 to age 65. It was assumed in the ERR

model that relative risk was the same for the US population and Japanese

Atom-bomb survivors.



ICRP publication 60 (ICRP, 1991) recommended the estimation of all cancer risks by

United Nations Scientific Committee on the Effects of Atomic Radiation 1988 report

(UNSCEAR, 1988), which were obtained by applying a model developed from

Atom-bomb survivor mortality data for the period 1950-1985 to the 1982 Japanese

population data. The lifetime risk was expressed as the risk of death associated with

exposure. The recommended ERR was a function of exposed age remaining constant

from 10 years after exposure to the end of life. For 0.2 Gray or below 0.1 Gray per

hour (low dose or low dose rate) the risks were acquired by dividing the high-dose

data by a DDREF of 2. The ICRP publication 60 also recommended organ factors

which can be used to estimate the organ-specific mortality risk by multiplying all

cancer risk by the factors. The factors are useful for a world population, assuming

that relative risks are constant over time.



National Council on Radiation Protection & Measurements report 1993 (NCRP, 1993)

reviewed the risk models in United Nations Scientific Committee on the Effects of

Atomic Radiation Report 1988 and ICRP publication 60, and supported the


                                              34
CHAPTER 2                                                                    BS HUANG


recommendations in ICRP publication 60.



US Environmental Protection Agency (EPA, 1994) used the geometric average of

EAR model and ERR model to calculate the solid cancer risks for the US population

except for breast cancer, for which the model in BEIR V report was used.



United Nations Scientific Committee on the Effects of Atomic Radiation 2000b

report developed mortality and incidence models for all solid organ cancers based on

Atom-bomb survivor mortality data for the period 1950-1990, and incidence data for

1958-1987 using ERR model and EAR model, without recommending any

preference between these two models. The report also estimated the lifetime risks for

populations of China, Japan, Puerto Rico, UK, and US.


2.5.1.2 Cancer risk estimating method from BEIR VII report

Based on the life span study (LSS) data of Atom-bomb survivors over the period

1958-1998 and data from medical and occupational related radiation studies, the

Biological Effects of Ionizing Radiation 7th Report (BEIR VII Phase 2) (NRC BEIR

Committee, 2006) published in 2006 provides an up-to-date method to estimate the

risk of cancer incidence from low-level ionizing radiation. The data for lifetime risks

are calculated for all solid cancers, as a function of sex, exposed age, radiation dose,

and time since exposure for the US population.



Both the ERR and EAR models were applied for estimating the excess cancer

incidence and mortality in this report. The ERR(D,e,a) and EAR(D,e,a) models for

stomach, colon, liver, lung, female breast, prostate, uterus, ovary, bladder, and all



                                           35
CHAPTER 2                                                                     BS HUANG


other solid cancers were set up independently from analyzing data of LSS cohort

(1958~1998) and other medical and occupational cohorts. The lifetime risk estimated

in this report is in the form of lifetime attributable risk (LAR), which is the defined

as the sum of every year’s cancer probability after exposure. From these two models,

two LARs were obtained. To calculate the risk for the US population, a weighted

LAR is calculated by weighting EAR models and ERR models (for most organs, the

weights are 0.3 for EAR and 0.7 for ERR) on a logarithmic scale. To estimate the

risks from exposure to low doses and low dose rates, a DDREF of 1.5, as

recommended in this report for all solid cancers was used. Thus the final LAR result

was acquired by dividing the weighted LAR( D, e) by 1.5. The LAR of cancer

incidence in the BEIR VII report for the US population is shown in Table 2.5.



The BEIR VII report supports the linear no-threshold (LNT) risk model for low dose

radiation transfer such as medical X-rays. According to this assumption the cancer

risk can be estimated in a linear fashion with no lower dose threshold. According to

the BEIR VII report, there is no evidence to show that this method is not applicable

for internal radiation dose, although there may be some uncertainty. In our studies,

we estimated the induced cancer risk of PET dose using the above same method

treating it like the CT dose.



In our research, for calculating the LAR for the HK and the US population, a table

like the Table 2.5 which was set up for the US population in 2006 was set up for the

HK population and this table was updated using newer data for the US population

(refer to details in Chapter 3).




                                           36
CHAPTER 2                                                                   BS HUANG


2.5.2 Organ-specific risk estimation

The LAR table of cancer incidence can be applied for calculating cancer risk for

different genders and ages, and for low dose radiation such as medical X-rays

(applying the LNT risk model). Linear interpolation is also applied for the ages for

which the risk data are not provided in the LAR table from the two nearest tabulated

ages. For example, if the liver dose measured for 25-year-old males is 50 mSv and

the LAR of liver cancer incidence for 25-year-old males for 100 mSv liver dose is 26

cases (derived from (30+22)/2) per 100 000 according to Table 2.5, then the LAR of

liver cancer incidence is (50/100)*(26/100 000) or 0.013%. In this way all the organ

specific LARs of cancer incidence for a specific age and gender can be calculated

from all the organ doses. For the risk of “other solid” cancer, a composite radiation

dose is appointed by weighting each organ using its corresponding tissue weighting

factor recommended in the ICRP publication 103, and this is used to calculate the

cancer risk.


2.5.3 Whole-body risk estimation

LARs of cancer incidence for all the solid cancers were summed to calculate the

whole body cancer incidence. Risk for males and females were calculated

independently. An example of organ-specific and whole body cancer risk calculation

is shown in Table 2.6 and 2.7. The organ doses were assumed to be 10 mSv and risk

data in Table 2.5 were applied. In Table 2.6, LARs of organs for which risk data are

listed in Table 2.5 were calculated. The other organ doses (assumed to be 10 mSv)

and corresponding WT are shown in Table 2.7, and the sum of WT was 0.307.



The composite dose were calculated by

                                          37
CHAPTER 2                                                                     BS HUANG


                              Dothersolid = ∑(WT ∗HT / 0.307)
                                         T


Dothersolid is the composite dose, H T stands for organ-specific dose, WT is the tissue

weighting factor for an specific organ or tissue T as listed in ICRP publication 103.

According to the organ doses shown in Table 2.7, the composite dose is 10 mSv. The

LAR data of “other solid” cancer for 20-year-old females is 323 (Table 2.5). Hence

the LAR for “other solid” cancer is

                     10/100*323=32.3 cases in 100,000 people

The whole body cancer risk is obtained by summing the estimated LAR in Table 2.6

and 32.3, and is equal to 157.6. This means that the LAR of cancer incidence for

20-year-old females in USA who receive a uniform dose of 10 mSv is 0.158%

(157.6/100000).


2.6 Comparison with baseline lifetime cancer incidence

To study how severe the estimated cancer risk associated with CT or PET-CT scan is ,

the LAR caused by the radiation and the baseline lifetime cancer incidence were

compared, and the proportion that LAR contributed to the total lifetime cancer

incidence was calculated. To do this, the baseline and the total lifetime risk of cancer

incidence were calculated.


2.6.1 Baseline lifetime risk of cancer incidence in the HK population

In our study the baseline lifetime cancer incidence was defined as the total cancer

incidence in one’s lifetime until death. Using the Hong Kong Cancer Statistics 2005

(Hong Kong Cancer Registry, 2005) the baseline lifetime cancer incidence was

calculated. Table 2.8 shows how we calculated the baseline lifetime risks, which is

defined as the probability of having cancer for a person in his or her lifetime (a life

                                             38
CHAPTER 2                                                                    BS HUANG


expectancy of 80 years). We assumed that 100000 infants were born in one year.

Since from 0 to 4 years old there are 104.96 cancer cases (Hong Kong Cancer

Registry, 2005), the number of healthy children at age 5 would be 999895.04

(100000-104.96). From 5 to 9 years old there are 46.44 cancer cases (999895.04

multiplied by the cancer incidence at this 5-9 years age group), therefore the number

of healthy people at age 10 would be 99848.60 (999895.04-46.44) etc. In this way, at

the age of 80, the total cancer cases are 32937.32, and so the baseline lifetime risk of

cancer incidence for 0 year old babies is 32.94%. We also calculated the baseline

lifetime risk of cancer incidence for people of all other ages up to age 70, as shown

in Table 2.9.


2.6.2 Baseline lifetime risk of cancer incidence calculation for the US
population

The baseline lifetime risk of cancer incidence for the US population has been

calculated by the US National Cancer Institute, as shown in its report Cancer

Statistics Review, 1975-2005 (US National Cancer Institute, 2006a) The results as

summarized in Table 2.10.


2.6.3 Comparison between baseline lifetime risk and LAR of cancer
incidence

The total lifetime cancer incidence was obtained by adding LAR to the baseline

lifetime cancer incidence, and the proportion was calculated by dividing the LAR

with total lifetime cancer incidence.




                                           39
CHAPTER 2                                                                BS HUANG


2.7 Discussion of the limitations in methodology

There are inherent errors and limitations to the methodology of our study. The

readings of the TLDs (TLD-100 or TLD-100H) scanned by uniform X-ray exposure

had an error of ±5%. The directionality error associated with the edge and surface of

TLDs was found to be ±2% (Wagner et al, 2000). Therefore the readings of TLD

would have an error of ±7%. Besides the TLD reading error, an estimated 1% error

in TLD calibration procedure also compounds the error of CT dose measurement.

Positioning of the TLDs would introduce an error to the CT dose results as radiation

exposure varies with locations. Although we used as many TLDs as possible to

minimize the variation, positioning uncertainty is unavoidable.



According to Groves et al, the doses measured directly using TLD were 18% higher

than the computer simulated doses using ImPACT (Groves et al, 2004). The

underestimation by the ImPACT spreadsheet may be attributed to differences

between the Rando phantom (used in the work of Groves and in our study) and the

MIRD mathematical phantoms (used in ImPACT). Although the dose measurement

using TLD is more accurate, it is time-consuming, tedious, and cannot be carried out

without equipment such as TLD reader, TLD oven, etc, while estimating the dose

using ImPACT is straightforward.



Uncertainties of PET dose are due to inherent error of the dose coefficients used for

PET dose estimation, which are calculated based on kinetic data of FDG in the

human body which in itself are associated with errors. Large variations observed in

the dose coefficients from the two publications (Table 2.2 and 2.3) highlight these

uncertainties.

                                         40
CHAPTER 2                                                                      BS HUANG




The uncertainties of cancer risk estimation are firstly produced from estimation using

the BEIR VII method which is based mainly on external radiation exposure, although

there is no conclusive data to show that the risk from internal exposure (such as

radiation in PET scan) would differ (NRC BEIR Committee, 2006). There is

additional uncertainty in the assumptions for establishing the LAR table of cancer for

the US population, and more so for the HK population (refer to the detailed

discussion in section 3.4.2, Chapter 3).



Although some other methods have been recommended for estimating radiation

induced cancer risks (ICRP, 1991; National Institutes of Health, 2003; NRC BEIR

Committee, 1990; UNSCEAR, 2000b; Upton AC et al, 2001), we used the method of

BEIR VII report as it is based on review of current research and the most updated

data, such as the data of Japanese Atom-bomb survivors. However, this method is

still controversial for cancer risk estimation (refer to section 3.4.3, Chapter 3).


2.8 Summary and conclusion

The main methods for estimating radiation doses and cancer risks applied in our

studies have been introduced in this chapter. Humanoid phantoms and TLDs were

used to measure the organ doses from CT scan directly. CT dose was also simulated

using a CT patient dosimetry calculator — ImPACT. The method introduced in ICRP

publication 80 was selected to estimate the internal dose of PET scan. Effective dose

were calculated according to the tissue weighting factors recommended in ICRP

publication 103. To estimate the lifetime risk of cancer incidence for the US and the

HK population, the method introduced in BEIR VII report was used.


                                            41
CHAPTER 2                                  BS HUANG



Figures:




            Figure 2.1:    Rando phantom




                          42
CHAPTER 2                                                    BS HUANG




      Figure 2.2:   Thermoluminescent dosimeter chips (TLD-100)




                                 43
CHAPTER 2                                     BS HUANG




            Figure 2.3:   Pediatric phantom




                          44
CHAPTER 2                                                BS HUANG



                            TLD Selection



                           TLD Calibration



                             TLD Pre-heat



                    TLD Placement (into phantom)



                               CT Scan



                   TLD Readout (24 hours after scan)




Figure 2.4:   Process of CT dose measurement using TLD




                                     45
CHAPTER 2                                                    BS HUANG




Figure 2.5:   Annealing oven for pre-heating TLD chips. (PTW-Freiburg,
              Freiburg, Germany)




                                   46
CHAPTER 2                                                   BS HUANG




Figure 2.6:   TLD reader (Harshaw, model QS5500, OH, USA)




                                 47
CHAPTER 2                                                            BS HUANG




Figure 2.7:   CT dose calculation using ImPACT spreadsheet. The parameters,
              including the CT scanner type, protocol parameters, and the CTDI
              values are input in the two above tables. The organ doses are then
              calculated in the two below tables, and then the effective dose is
              calculated based on the organ doses and tissue weighting factors
              recommended in ICRP publication 60.




                                      48
CHAPTER 2                                                              BS HUANG



Tables:



  charge (nC)    air kerma (mR)       tube current (mA)   tube potential (kV)
  26.99          318.2                10                  120
  239.22         3155                 100                 120
  485.40         6314                 200                 120
  742.45         9463                 300                 120
  943.65         12890                400                 120
  32.22          425.4                10                  140
  332.57         4268                 100                 140
  639.99         8532                 200                 140
  991.51         12780                300                 140
  1348.47        17390                400                 140

Table 2.1:   Thermoluminescent dosimeters (TLD-100) calibration. Charge is
             from the reading from TLD reader; air kerma is from the reading
             of ionization chamber; tube current and tube potential are the
             settings of CT scanner.


                Key:

                nC = nano coulomb,

                mR = milliroentgen.




                                        49
CHAPTER 2                                                         BS HUANG



                     organ               FT (mGy/MBq)
                     gonads              0.014
                     lung                0.01
                     stomach             0.011
                     colon               0.013
                     bone marrow         0.011
                     esophagus           0.011
                     thyroid             0.01
                     liver               0.011
                     bladder             0.16
                     breast              0.007
                     bone surface        0.011
                     skin                0.008
                     brain               0.011
                     thymus              0.011
                     spleen              0.011
                     adrenal             0.011
                     pancreas            0.011
                     kidney              0.011
                     large intestines    0.011
                     small intestines    0.011
                     uterus              0.021
                     muscle              0.011
                     heart               0.011

Table 2.2:   Organ-specific dose coefficients (FT) for calculating the organ
             doses of 18F-FDG PET scan, recommended in ICRP publication 80
             (ICRP, 1998).




                                        50
CHAPTER 2                                                             BS HUANG




                      organ                F (mGy/MBq)
                      gonads               0.011
                      lung                 0.015
                      bone marrow          0.011
                      liver                0.024
                      bladder              0.073
                      brain                0.046
                      spleen               0.015
                      pancreas             0.014
                      kidney               0.021
                      heart                0.068

Table 2.3:   Organ-specific dose coefficients (F) for calculating the organ doses
             of 18F-FDG PET scan, recommended in MIRD report (Hays et al,
             2002).




                                      51
CHAPTER 2                                                            BS HUANG




                                      tissue weighting factor (WT)
             organ              ICRP 60             ICRP 103
             gonads             0.2                 0.08
             lung               0.12                0.12
             stomach            0.12                0.12
             colon              0.12                0.12
             bone marrow        0.12                0.12
             esophagus          0.05                0.04
             thyroid            0.05                0.04
             liver              0.05                0.04
             bladder            0.05                0.04
             breast             0.05                0.12
             bone surface       0.01                0.01
             skin               0.01                0.01
             brain              0.005               0.01
             thymus             0.005               0.013
             spleen             0.005               0.013
             adrenal            0.005               0.013
             pancreas           0.005               0.013
             kidney             0.005               0.013
             large intestines   0.005               0.013
             small intestines   0.005               0.013
             uterus             0.005               0.013
             muscle             0.005               0.013
             heart              0                   0.013

Table 2.4:      Tissue weighting factors (WT) recommended in ICRP publication
                60 (ICRP, 1991) and publication 103 (ICRP, 2007).




                                          52
CHAPTER 2                                                             BS HUANG




                                    age at exposure(years)
cancer site   0        10     20      30      40     50    60       70     80
  Males
stomach       76       55     40      28     27      25     20      14     7
colon         336      241    173     125    122     113    94      65     30
liver         61       43     30      22     21      19     14      8      3
lung          314      216    149     105    104     101    89      65     34
prostate      93       67     48      35     35      33     26      14     5
bladder       209      150    108     79     79      76     66      47     23
other solid   1123     503    312     198    172     140    98      57     23
thyroid       115      50     21      9      3       1      0.3     0.1    0
              2563     1445   977     686    648     591    489     343    174
  Females
stomach
colon         101      72     52      36     35      32     27      19     11
liver         220      158    114     82     79      73     62      45     23
lung          28       20     14      10     10      9      7       5      2
breast        733      504    346     242    240     230    201     147    77
uterus        1171     712    429     253    141     70     31      12     4
ovary         50       36     26      18     16      13     9       5      2
bladder       104      73     50      34     31      25     18      11     5
other solid   212      152    109     79     78      74     64      47     24
thyroid       1339     523    323     207    181     148    109     68     30

Table 2.5:    Table of LAR of cancer incidence for the US population. The data
              in the table is the numbers of cases per 100,000 persons exposed to
              a radiation dose of 100 mGy. The table is from BEIR VII report
              (NRC BEIR Committee, 2006).




                                       53
CHAPTER 2                                                          BS HUANG




        organ       organ dose/mSv     LAR for 100mSv   estimated LAR
        ovary       10                 50               5
        colon       10                 114              11.4
        stomach     10                 52               5.2
        lung        10                 346              34.6
        thyroid     10                 113              11.3
        liver       10                 14               1.4
        bladder     10                 109              10.9
        breast      10                 429              42.9
        uterus      10                 26               2.6

Table 2.6:   Examples of organ-specific LAR calculation (for 20-year-old
             females). LAR for 100 mSv is the numbers of cancer cases per 100,
             000 people (refer to Table 2.5); estimated LAR is calculated by
             linear extrapolation (for example, for ovary, it is derived by
             50*10/100).


                  Keys:
                  LAR= lifetime attributable risk.




                                         54
CHAPTER 2                                                        BS HUANG




              organ               organ dose/mSv    WT
              bone marrow         10                0.12
              bone surface        10                0.01
              skin                10                0.01
              brain               10                0.01
              thymus              10                0.013
              kidney              10                0.013
              pancreas            10                0.013
              spleen              10                0.013
              adrenal             10                0.013
              esophagus           10                0.04
              small intestines    10                0.013
              large intestines    10                0.013
              muscle              10                0.013
              heart               10                0.013

Table 2.7:   Example of organ doses for calculating the composite dose (for
             females). The tissue weighting factors (WT) are from ICRP
             publication 103.




                                    55
CHAPTER 2                                                              BS HUANG




       years      cancer cases     healthy people     total cancer cases
       0-         104.96           100000.00          104.96
       5-         46.44            99895.04           151.40
       10-        49.42            99848.60           200.81
       15-        90.78            99799.19           291.60
       20-        107.64           99708.40           399.24
       25-        144.34           99600.76           543.57
       30-        244.42           99456.43           788.00
       35-        409.07           99212.00           1197.07
       40-        694.60           98802.93           1891.66
       45-        1085.15          98108.34           2976.81
       50-        1848.59          97023.19           4825.41
       55-        2536.99          95174.59           7362.40
       60-        4113.09          92637.60           11475.49
       65-        5849.34          88524.51           17324.83
       70-        6994.52          82675.17           24319.35
       75-        8617.97          75680.65           32937.32

Table 2.8:     Example of baseline lifetime risk of cancer incidence calculation.
               “healthy people” means the number of healthy people at the
               beginning of these years. “cancer cases” means the number of
               cancer cases for these people during these years. “total cancer
               cases” is acquired by summing all the previous cancer cases. For
               calculation, it is assumed that at the beginning there are 100,000
               healthy 0-year-old babies.




                                        56
CHAPTER 2                                                            BS HUANG



                     age (years)     males      females
                     0               32.94%     23.37%
                     5               32.87%     23.32%
                     10              32.84%     23.29%
                     20              32.74%     23.19%
                     30              32.57%     22.95%
                     40              32.12%     22.18%
                     50              30.88%     20.10%
                     60              27.61%     16.50%
                     70              18.88%     10.59%

Table 2.9:   Baseline lifetime risk of cancer incidence calculated for males and
             females of different ages in Hong Kong.




                                      57
CHAPTER 2                                                             BS HUANG



                       age (years)   males      females
                       0             43.89%     37.35%
                       10            44.37%     37.66%
                       20            44.50%     37.64%
                       30            44.86%     37.50%
                       40            45.14%     36.92%
                       50            45.33%     35.35%
                       60            43.84%     32.08%
                       70            37.74%     26.17%

Table 2.10:   Baseline lifetime risk of cancer incidence for males and females of
              different ages in US (US National Cancer Institute, 2006a).




                                       58
CHAPTER 3                                                                   BS HUANG




CHAPTER 3              CANCER RISK ESTIMATING TABLES

FOR THE HONG KONG AND UNITED STATES

POPULATION


3.1 Introduction

In order to evaluate the cancer risk associated with CT and PET-CT radiation

exposure to the United States (US) population and the Hong Kong (HK) population,

the specific cancer risk tables for the HK and the US populations were established

(HK) or updated (US). The tables are for an organ dose of 100 mSv, and can be used

to calculate the cancer risks for specific doses and organs applying the linear

no-threshold (LNT) model. The data of cancer statistics and life table of Hong Kong

were used for establishing this table. We also aimed to update the table for the US

population in the BEIR VII report, using the updated data of US Cancer Statistics

2001-2005 (US National Cancer Institute, 2006b), and US Life Table 2005 (US

Disaster Center, 2006). These tables set up or updated can be used not only for

medical radiation, but also can be applied for the other low-dose ionizing radiation.




                                          59
CHAPTER 3                                                                      BS HUANG



3.2 Methodology

The method introduced in BEIR VII report (NRC BEIR Committee, 2006) was

applied to estimate the radiation induced cancer risk in the form of lifetime

attributable risk (LAR), which is calculated by

                          LAR( D, e) = ∑ M ( D, e, a) S (a) / S (e)
                                         a


Where D is dose (assumed to be 100 mSv, for calculating the table), e is exposed age,

a is attained age which is from e+L to 100 (L is a risk-free latent period which equals

5) accounting for “lifetime”. S(a), S(e) is the probability of surviving until age a and

e respectively, which is obtained in the life table data. Hence the ratio of S(a) / S(e) is

the probability of surviving to age a on condition that he/she survives to age e.

M(D,e,a) is the excessive cancer risk in a specific year, which can be calculated

using ERR model and EAR model. In our study, the cancer risk estimating table was

set up by using both EAR and ERR models for males and females with a series of

ages (0, 10, 20,…, 80) when the patients are exposed to radiation.


3.2.1 ERR model

Using ERR model, M(D,e,a) is

                              M ( D, ea) = ERR( D, e, a)λc (a)
                                                         I                      (3-1)


The λ I (a) represents sex- and age-specific baseline cancer incidence from cancer
     c




statistics report, which is expressed as cases in 100,000 persons in a year, and

ERR(D,e,a) is the excess relative risk. And then the LAR was calculated by

               LAR( D, e) = ∑ ERR( D, e, a )λc (a) S (a) / S (e)
                                             I                        (3-2)
                              a


For sites other than breast and thyroid, ERR(D,e,a) is given by


                                             60
CHAPTER 3                                                                        BS HUANG


                 ERR( D, e, a ) = β s D exp(γe ∗ )(a / 60)η            (3-3)

For female breast, ERR is given by

                          ERR / Sv = β (a / 60) −2                     (3-4)

Where β s (or β) is organ- and gender- specific factor (NRC BEIR Committee,

                                                           ∗
2006), D is the dose (Sv), e is age at exposure in years, e is equal to (e-30)/10 when

e<30, and equal to zero when e>30 or e=30, and a is attained age in years. γ and

η is given by BEIR VII report (NRC BEIR Committee, 2006). For thyroid cancer

the model is a little different (η =0), and only the ERR model is suggested in the

report. The equation for estimating the LAR of thyroid cancer is the same as “all

solids” cancers (3-2).


3.2.2 EAR model

For EAR model, M(D,e,a) equals to EAR(D,e,a), then LAR is given by

                 LAR ( D, e) = ∑ EAR( D, e, a ) S (a) / S (e)         (3-5)
                                  a


for sites other than breast and thyroid, EAR(D,e,a) is also of the form as equation

(3-5), with different β s , γ and η from ERR model, as recommended in BEIR

VII report.

                         EAR( D, e, a ) = β s D exp(γe ∗ )(a / 60)η      (3-6)

for female breast, EAR is given by

                           EAR per 104 woman-years per gray=

                               9.4 exp[(-0.05(e - 30))](a/60)η        (3-7)

There is no EAR model for thyroid cancer suggested in BEIR VII report.




                                               61
CHAPTER 3                                                                     BS HUANG


3.2.3 Method for setting up the cancer risk estimating table for the HK
population

For calculating the LAR from the dose data for the HK population, a table like the

Table 2.5 should be established following the principles and assumptions in BEIR

VII report. Also to be taken into consideration are the Hong Kong cancer statistics

data (Hong Kong Cancer Registry, 2005) and the Hong Kong life table (HK Census

and Statistics Department, 2006). The table is specific for gender, age, organ, and

100 mSv absorbed dose. In our research, we calculated the weighted LAR for the HK

population by weighting the two results (EAR and ERR models) using the same

weights as what were used in the BEIR VII report for the US population, as there

have not been relevant studies on the risk transport from Japanese to the HK

population and no weights were recommended for other populations in the BEIR VII

report except the US population. The weights for the EAR and ERR models were

somewhat different for the organs; for sites other than breast, thyroid, and lung, a

weight of 0.7 is used for the estimate obtained using relative risk transport (that is,

using ERR model) and a weight of 0.3 for the estimate from absolute risk transport

(using EAR model), with the weighting done on a logarithmic scale. For lung the

weights are reversed (0.3 for ERR and 0.7 for EAR). Only ERR model was

suggested for thyroid cancer, hence LAR of thyroid cancer were only related to the

ERR. For female breast, the BEIR VII report preferred the estimates based on EAR

model. To sum up, the averaged LAR acquired from results of EAR and ERR model

were, for all organs (except breast, lung and thyroid)

               LAR = (exp(0.3 * ln(EAR) + 0.7 * ln(EAR)))/1.5         (3-8)

for thyroid,

                                    LAR = ERR/1.5

                                           62
CHAPTER 3                                                                    BS HUANG


for female breast,

                                    LAR = EAR/1.5


3.2.4 Method for updating the cancer risk estimating table for the US
population

To calculate the LAR of cancer incidence for the US population, the table in BEIR

VII report was re-calculated and updated using the same methods as used for

establishing the table for the HK population, the US Cancer Statistics 2001-2005 (US

National Cancer Institute, 2006b) and US Life Table 2005 (US Disaster Center,

2006).


3.3 Results

3.3.1 ERR for HK males and females

According to Equation 3-1~3-4, the LAR table for HK males and females were

calculated using the HK life table and cancer statistics data (HK Census and

Statistics Department, 2006; Hong Kong Cancer Registry, 2005), for each typical

organ (Table 3.3). Firstly, ERR(D,e,a) is calculated for a specific year a and a specific

exposed year e according to Equation 3-3 and 3-4; then the ERR(D,e,a) was

multiplied with baseline cancer incidence to calculate the excess cancer incidence.

The products weighted by S(a)/S(e) for each year a (a is from e+L to 100) were

summed to calculate the LAR (Equation 3-2). This process was repeated for a series

of exposed ages and organs. Table 3.1 (male) and 3.2 (female) shows an example of

our calculated table for estimating cancer risk of the HK population using ERR

model. Due to the limits of page width and content, only the calculation procedure of

stomach, colon, liver of 80-year-old males and females are shown for ERR model.

                                           63
CHAPTER 3                                                                BS HUANG


Table 3.3 shows the summary of LAR calculated by using ERR model.


3.3.2 EAR for HK males and females

For EAR model, the calculation procedure for all the related cancers (stomach, colon,

liver, lung, uterus, ovary, bladder, breast, other solid) of 80-year-old males and

females are shown in Table 3.4 and 3.5 respectively. Firstly the EAR(D,e,a) for a

specific year a and a specific exposed year e ware calculated according to Equation

3-6 and 3-7, then results weighted by S(a)/S(e) were summed to calculate the LAR

(Equation 3-5). Table 3.6 shows the summary of LAR calculated by using EAR

model. Repetition for a series of exposed ages and organs were done and the LAR

table was set up (Table 3.6).


3.3.3 Cancer risk estimating table for the HK population

The organ-specific LARs of cancer incidence for the HK population of different ages

calculated according to EAR and ERR models were averaged with the weights

introduced above. The newly established table for the HK population is shown in

Table 3.7.


3.3.4 Updated cancer risk estimating table for the US population

By the same means as for HK population, the LAR tables for US population were

also established using data of year 2005 (US Disaster Center, 2006; US National

Cancer Institute, 2006a) and were averaged with the weights discussed above. The

updated table is shown in Table 3.8.




                                         64
CHAPTER 3                                                                 BS HUANG


3.4 Discussion

3.4.1 Comparison between cancer risk estimating tables for the HK and
the US population

Comparing the two LAR tables for the HK and the US population (Table 3.7 and 3.8),

the cancer risks data are different in the following aspects. The LAR for cancer

incidence in the colon, uterus, ovary, and bladder in HK females are lower than in

US females. For males, only LAR of prostate and bladder cancer are lower in the HK

population than in the US population. LARs for other organ cancers except thyroid

are higher in the HK population than in the US population. For thyroid, the cancer

risk is lower in the HK population than in the US population at young ages (below

about 35 years) and higher at old ages (above 40 years). The difference between

these two tables is due to the dissimilarity in the baseline cancer incidence between

these two populations, as the methods used to set up these two tables are the same.


3.4.2 Uncertainties of setting up the cancer risk estimating tables using
BEIR VII report

There are several uncertainties in setting up the LAR tables using the principles in

the BEIR VII report. Firstly, uncertainties are produced from the developing ERR or

EAR models from LSS data of Atom-bomb survivors (Equation 3-3 ~ 3-7). Secondly,

we used a DDREF of 1.5 to calculate the low-dose radiation risk based on risk tables

from LSS data. However this is not supported by robust research data (there are also

some values of DDREF recommended in other publications) and produces

uncertainties. Thirdly, errors are incurred by using the weights used for EAR and

ERR models for radiation induced risks transport from a Japanese population to other


                                         65
CHAPTER 3                                                                      BS HUANG


populations. There are additional errors from the use of the same weights for EAR

and ERR for the risk transport between the Japanese and the HK population as this

may differ from between the Japanese and the US population. Our study is limited by

the fact that the risk transport analysis, which is based on complex analyses including

comparisons of the factors that increase cancer rate and the effect of radiation on

these factors in both populations, has not been studied for the HK population.


3.4.3 Controversy of cancer risk estimation using BEIR VII report

Risk estimation according to the method in BEIR VII report is controversial in that it

may be overestimated. In the calculation of the lifetime risk, this method sums the

risks from the age at 5 years after exposure to the age of 100 years old. However, the

average natural life span in both US is currently less than 80 years. Also, the

assumptions in BEIR VII report for risk of ionizing radiation with low dose (or low

dose-rate) remain controversial. Firstly, whether cancer is itself a risk of low dose

radiation especially dose from CT scan is still unproven as there have not been any

epidemiologic studies to date to support this (Brenner et al, 2007; Tubiana et al,

2008). However, there is direct epidemiologic data from 30,000 Atom-bomb

survivors who were on the peripheries of Hiroshima and Nagasaki exposed to the

same low dose range as CT scan (5-100 mSv) showing a small but statistically

significant increase in the risk of cancer (Pierce et al, 2000). In addition, a significant

excess risk was reported in a cohort of 400,000 workers in the nuclear industry who

were exposed to an average dose of approximately 20 mSv (Cardis et al, 2005).

Secondly, there is no experimental data to support the LNT extrapolation for low

dose risk estimation. However the majority of organizations have advised that LNT

best fits the data (United Nations Scientific Committee on the Effects of Atomic


                                            66
CHAPTER 3                                                                  BS HUANG


Radiation, 2000; Upton AC et al, 2001; Valentin J., 2005), and it is also advised that

linear extrapolation of cancer risk from low dose should be the most appropriate

method (Brenner et al, 2003).


3.5 Summary and conclusion

In this chapter, we introduced our newly developed tables for estimating the cancer

risk for US and the HK populations from radiation dose by using the methods

introduced in BEIR VII report, the cancer statistics data and the population data.

Although uncertainties exist in our table, we believe that the method used for

establishing the tables are the most appropriate method available (refer to Chapter 2

for other methods). These established tables can be used not only to estimate the

risks from medical radiation exposure, but also for risks from other radiation sources.

Future studies may be focused on setting up or improving the mortality tables for the

HK population.




                                          67
CHAPTER 3        BS HUANG



Tables:




            68
 CHAPTER 3                                                                                                                 BS HUANG


male                                                  stomach                      colon                       liver
e    (e-30)/10      a     a/60   S(a)     S(a)/S(e)   ERR       baseline   risk    ERR     baseline   risk     ERR     baseline   risk
80   0              81    1.35   0.5061   1.0000      0.14      193.36     26.68   0.41    320.24     132.54   0.21    161.22     33.89
     0              82    1.37   0.4696   0.9279      0.13      202.70     25.51   0.38    326.00     123.06   0.19    156.20     29.95
     0              83    1.38   0.4324   0.8543      0.11      205.76     23.44   0.34    327.16     111.79   0.17    162.88     28.27
     0              84    1.40   0.3948   0.7801      0.10      208.82     21.36   0.31    328.32     100.74   0.16    169.56     26.43
     0              85    1.42   0.3573   0.7060      0.09      211.88     19.29   0.27    329.48     89.99    0.14    176.24     24.45
     0              86    1.43   0.3203   0.6329      0.08      214.94     17.26   0.24    330.64     79.64    0.12    182.92     22.38
     0              87    1.45   0.2841   0.5614      0.07      218.00     15.28   0.21    331.80     69.75    0.11    189.60     20.25
     0              88    1.47   0.2492   0.4923      0.06      218.00     13.18   0.18    331.80     60.19    0.09    189.60     17.47
     0              89    1.48   0.2158   0.4263      0.05      218.00     11.24   0.15    331.80     51.31    0.08    189.60     14.89
     0              90    1.50   0.1843   0.3641      0.04      218.00     9.45    0.13    331.80     43.15    0.07    189.60     12.52
     0              91    1.52   0.1551   0.3065      0.04      218.00     7.83    0.11    331.80     35.75    0.05    189.60     10.38
     0              92    1.53   0.1284   0.2537      0.03      218.00     6.39    0.09    331.80     29.16    0.04    189.60     8.46
     0              93    1.55   0.1045   0.2064      0.02      218.00     5.12    0.07    331.80     23.36    0.04    189.60     6.78
     0              94    1.57   0.0834   0.1647      0.02      218.00     4.02    0.06    331.80     18.36    0.03    189.60     5.33
     0              95    1.58   0.0651   0.1287      0.01      218.00     3.10    0.04    331.80     14.13    0.02    189.60     4.10
     0              96    1.60   0.0497   0.0982      0.01      218.00     2.33    0.03    331.80     10.63    0.02    189.60     3.09
     0              97    1.62   0.0370   0.0731      0.01      218.00     1.71    0.02    331.80     7.80     0.01    189.60     2.27
     0              98    1.63   0.0268   0.0530      0.01      218.00     1.22    0.02    331.80     5.57     0.01    189.60     1.62
     0              99    1.65   0.0189   0.0373      0.00      218.00     0.85    0.01    331.80     3.86     0.01    189.60     1.12
                                                      LAR =                9.90                       45.27                       13.07



 Table 3.1:      Example of LAR calculation using ERR model (for HK males). ERR is the excess relative risk, defined by the rate of
                 cancer in an exposed population divided by the disease in an unexposed population minus 1. LAR is acquired by
                 summing the risks for each attained age (a).


                  Keys:

                                                                  69
CHAPTER 3                                                                                                             BS HUANG


            e=exposed age (years), a=attained age (years, from e+3 to 99), S(a)=probability of surviving until age a, S(a)/S(e)
            =the probability of surviving to age a on condition that he/she survives to age e, risk=ERR × baseline.




                                                            70
CHAPTER 3                                                                                                                   BS HUANG


female                                                stomach                        colon                         liver
e      (e-30)/10   a    a/60     S(a)     S(a)/S(e)   ERR       baseline   risk      ERR      baseline   risk      ERR      baseline   risk
80     0           81   1.3500   0.7053   1.0000      0.3153    56.6       17.8480   0.2825   71         20.0567   0.2102   78.4       16.4816
       0           82   1.3667   0.6753   0.9574      0.2968    55         16.3216   0.2658   63         16.7482   0.1978   67         13.2551
       0           83   1.3833   0.6428   0.9114      0.2777    49.6       13.7758   0.2488   61         15.1773   0.1852   61.2       11.3317
       0           84   1.4000   0.6081   0.8621      0.2584    44.2       11.4199   0.2315   59         13.6558   0.1722   55.4       9.5424
       0           85   1.4167   0.5713   0.8100      0.2388    38.8       9.2640    0.2139   57         12.1918   0.1592   49.6       7.8951
       0           86   1.4333   0.5327   0.7553      0.2190    33.4       7.3154    0.1962   55         10.7915   0.1460   43.8       6.3955
       0           87   1.4500   0.4926   0.6985      0.1993    28         5.5799    0.1785   53         9.4618    0.1329   38         5.0485
       0           88   1.4667   0.4514   0.6399      0.1797    28         5.0312    0.1610   53         8.5313    0.1198   38         4.5520
       0           89   1.4833   0.4093   0.5804      0.1604    28         4.4912    0.1437   53         7.6156    0.1069   38         4.0634
       0           90   1.5000   0.3671   0.5204      0.1416    28         3.9650    0.1269   53         6.7234    0.0944   38         3.5874
       0           91   1.5167   0.3251   0.4610      0.1235    28         3.4581    0.1106   53         5.8638    0.0823   38         3.1287
       0           92   1.5333   0.2841   0.4028      0.1063    28         2.9758    0.0952   53         5.0460    0.0709   38         2.6924
       0           93   1.5500   0.2446   0.3468      0.0901    28         2.5233    0.0807   53         4.2787    0.0601   38         2.2830
       0           94   1.5667   0.2071   0.2937      0.0752    28         2.1053    0.0674   53         3.5699    0.0501   38         1.9048
       0           95   1.5833   0.1723   0.2443      0.0616    28         1.7258    0.0552   53         2.9263    0.0411   38         1.5614
       0           96   1.6000   0.1406   0.1993      0.0496    28         1.3875    0.0444   53         2.3528    0.0330   38         1.2554
       0           97   1.6167   0.1123   0.1592      0.0390    28         1.0923    0.0349   53         1.8521    0.0260   38         0.9882
       0           98   1.6333   0.0876   0.1243      0.0300    28         0.8402    0.0269   53         1.4248    0.0200   38         0.7602
       0           99   1.6500   0.0667   0.0945      0.0225    28         0.6304    0.0202   53         1.0689    0.0150   38         0.5703
                                                      LAR =                5.2385                        8.3699                        4.6686




                                                                  71
CHAPTER 3                                                                                                               BS HUANG


Table 3.2:   Example of LAR calculation using ERR model (for HK females). ERR is the excess relative risk, defined by the rate
             of cancer in an exposed population divided by the disease in an unexposed population minus 1. Risk is the cancer
             incidence caused by 100 mSv radiation in a specific year. LAR is acquired by summing the risks for each attained
             age (a).


              Keys:

              e=exposed age (years), a=attained age (years, from e+3 to 99), S(a)=probability of surviving until age a, S(a)/S(e)
              =the probability of surviving to age a on condition that he/she survives to age e, risk=ERR × baseline.




                                                              72
CHAPTER 3                                                                     BS HUANG



                                          age at exposure(years)
cancer site   0          10       20         30     40     50        60      70      80
males
stomach       114.4      84.9     62.9       46.2    44.9    41.1    33.5    23.2    9.9
colon         584.9      433.7    320.4      236.5   231.1   215.8   180.4   116.2   45.3
liver         306.0      225.1    165.1      120.1   110.2   90.2    61.9    32.8    13.1
lung          695.8      516.1    382.7      283.2   277.4   261.1   215.9   133.7   52.6
prostate      95.3       70.7     52.5       39.1    39.4    39.7    35.6    24.2    11.5
bladder       232.9      172.8    128.2      94.9    93.0    88.6    76.3    54.8    24.6
other solid   3416.9     838.7    529.1      347.3   295.8   232.5   161.0   86.7    31.9
thyroid       187.3      81.2     34.9       14.1    12.3    10.0    6.0     3.7     0.1

females
stomach       178.2      132.1    98.0       72.0    66.7    55.0    41.3    20.0    5.2
colon         213.3      155.3    114.3      84.1    75.4    67.1    52.2    27.0    8.4
liver         413.9      301.6    219.5      160.7   134.3   96.8    60.6    21.9    4.7
lung          1382.1     998.6    739.1      548.2   499.9   440.7   375.8   232.4   95.3
uterus        2.3        1.7      1.3        0.9     0.8     0.6     0.3     0.1     0.1
ovary         66.1       48.6     34.6       24.7    19.2    12.8    7.6     4.0     1.1
bladder       297.5      210.3    155.6      115.6   114.0   111.9   102.0   80.8    47.8
other solid   4280.0     1577.6   1052.0     721.8   579.3   430.0   334.9   255.9   131.8
thyroid       1103.3     478.4    199.3      81.0    65.4    48.7    32.2    17.8    2.8
breast        346.5      343.3    341.8      332.5   242.4   148.6   89.0    52.1    21.8

Table 3.3:        Summary of LAR results using ERR model (for the HK
                  population). The data in the table is the numbers of cancer cases
                  per 100,000 persons exposed to a radiation dose of 100mGy.


                   Keys:
                   LAR= lifetime attributable risk.




                                             73
CHAPTER 3                                                                                                         BS HUANG


males                                                                        EAR*S(a)/S(e) of each organ
e    (e-30)/10   a    a/60     S(a)     S(a)/S(e)   stomach    colon     liver   lung       prostate    bladder   other solid
80 0             80   1.3333   0.5415   1.0000      10.9654    7.1611    7.1560  10.2662 0.2462         6.7424    13.8746
     0           81   1.3500   0.5061   0.9346      10.6113    6.9298    7.0377  10.2353 0.2382         6.7893    13.4266
     0           82   1.3667   0.4696   0.8672      10.1902    6.6548    6.8671  10.1229 0.2288         6.7809    12.8938
     0           83   1.3833   0.4324   0.7985      9.7061     6.3387    6.6447  9.9266     0.2179      6.7142    12.2812
     0           84   1.4000   0.3948   0.7291      9.1649     5.9852    6.3726  9.6464     0.2057      6.5875    11.5964
     0           85   1.4167   0.3573   0.6599      8.5744     5.5996    6.0545  9.2849     0.1925      6.4010    10.8493
     0           86   1.4333   0.3203   0.5915      7.9420     5.1866    5.6939  8.8449     0.1783      6.1550    10.0491
     0           87   1.4500   0.2841   0.5247      7.2765     4.7520    5.2957  8.3318     0.1633      5.8517    9.2070
     0           88   1.4667   0.2492   0.4601      6.5882     4.3025    4.8666  7.7534     0.1479      5.4955    8.3361
     0           89   1.4833   0.2158   0.3984      5.8886     3.8456    4.4141  7.1206     0.1322      5.0928    7.4508
     0           90   1.5000   0.1843   0.3403      5.1899     3.3893    3.9474  6.4463     0.1165      4.6520    6.5669
     0           91   1.5167   0.1551   0.2864      4.5049     2.9420    3.4759  5.7459     0.1011      4.1833    5.7001
     0           92   1.5333   0.1284   0.2372      3.8460     2.5116    3.0099  5.0358     0.0863      3.6985    4.8663
     0           93   1.5500   0.1045   0.1929      3.2246     2.1059    2.5594  4.3332     0.0724      3.2101    4.0801
     0           94   1.5667   0.0834   0.1539      2.6510     1.7313    2.1336  3.6550     0.0595      2.7310    3.3544
     0           95   1.5833   0.0651   0.1202      2.1334     1.3932    1.7408  3.0170     0.0479      2.2734    2.6994
     0           96   1.6000   0.0497   0.0918      1.6773     1.0954    1.3874  2.4324     0.0377      1.8483    2.1223
     0           97   1.6167   0.0370   0.0684      1.2857     0.8396    1.0779  1.9115     0.0289      1.4646    1.6268
     0           98   1.6333   0.0268   0.0495      0.9587     0.6261    0.8145  1.4608     0.0215      1.1285    1.2131
     0           99   1.6500   0.0189   0.0348      0.6937     0.4530    0.5972  1.0831     0.0156      0.8436    0.8778
                                LAR      =          62.4349    40.7738   47.0688 76.4565 1.4016         55.0295   78.9993




                                                              74
CHAPTER 3                                                                                                             BS HUANG


Table 3.4:   Example of LAR calculation using EAR model (for HK males). EAR is the excess absolute risk, defined as the rate
             of cancer in an exposed population minus the rate of cancer in an unexposed population. Here it means the excess
             cancer incidence caused by 100 mSv radiation in a specific year. LAR is acquired by summing the EAR for each
             attained age (a).


              Keys:

              e=exposed age (years), a=attained age (years, from e+3 to 99), S(a)=probability of surviving until age a, S(a)/S(e)
              =the probability of surviving to age a on condition that he/she survives to age e, risk=EAR.




                                                              75
   CHAPTER 3                                                                                                               BS HUANG


females                                                                         EAR*S(a)/S(e) of each organ
e (e-30)/10    a    a/60     S(a)     S(a)/S(e)   stomach   colon    liver    lung    uterus ovary       bladder   other solid   breast
80 0           80   1.3333   0.7329   1.0000      10.9654   3.5805   3.2527   15.1761 2.6854 1.5665 4.2140         10.7416       12.8991
    0          81   1.3500   0.7053   0.9623      10.9259   3.5676   3.2938   15.5790 2.6757 1.5608 4.3691         10.7029       12.5841
    0          82   1.3667   0.6753   0.9213      10.8259   3.5350   3.3161   15.8978 2.6512 1.5466 4.5025         10.6050       12.2115
    0          83   1.3833   0.6428   0.8770      10.6613   3.4812   3.3176   16.1182 2.6109 1.5230 4.6094         10.4437       11.7806
    0          84   1.4000   0.6081   0.8297      10.4294   3.4055   3.2963   16.2274 2.5541 1.4899 4.6853         10.2165       11.2920
    0          85   1.4167   0.5713   0.7795      10.1291   3.3075   3.2510   16.2142 2.4806 1.4470 4.7260         9.9224        10.7485
    0          86   1.4333   0.5327   0.7269      9.7596    3.1868   3.1804   16.0675 2.3901 1.3942 4.7273         9.5605        10.1525
    0          87   1.4500   0.4926   0.6722      9.3217    3.0438   3.0838   15.7784 2.2829 1.3317 4.6853         9.1315        9.5083
    0          88   1.4667   0.4514   0.6158      8.8183    2.8794   2.9609   15.3413 2.1596 1.2598 4.5974         8.6383        8.8217
    0          89   1.4833   0.4093   0.5585      8.2544    2.6953   2.8125   14.7550 2.0215 1.1792 4.4618         8.0859        8.1005
    0          90   1.5000   0.3671   0.5008      7.6375    2.4939   2.6404   14.0234 1.8704 1.0911 4.2787         7.4816        7.3541
    0          91   1.5167   0.3251   0.4436      6.9774    2.2783   2.4471   13.1557 1.7088 0.9968 4.0496         6.8350        6.5935
    0          92   1.5333   0.2841   0.3876      6.2863    2.0527   2.2363   12.1677 1.5395 0.8980 3.7783         6.1580        5.8310
    0          93   1.5500   0.2446   0.3337      5.5780    1.8214   2.0124   11.0805 1.3661 0.7969 3.4706         5.4642        5.0798
    0          94   1.5667   0.2071   0.2826      4.8679    1.5895   1.7808   9.9212  1.1921 0.6954 3.1342         4.7685        4.3532
    0          95   1.5833   0.1723   0.2351      4.1716    1.3622   1.5472   8.7209  1.0216 0.5959 2.7785         4.0865        3.6641
    0          96   1.6000   0.1406   0.1918      3.5048    1.1444   1.3177   7.5134  0.8583 0.5007 2.4139         3.4333        3.0241
    0          97   1.6167   0.1123   0.1532      2.8817    0.9410   1.0982   6.3332  0.7057 0.4117 2.0516         2.8229        2.4430
    0          98   1.6333   0.0876   0.1196      2.3144    0.7557   0.8938   5.2132  0.5668 0.3306 1.7027         2.2672        1.9282
    0          99   1.6500   0.0667   0.0910      1.8119    0.5916   0.7090   4.1820  0.4437 0.2588 1.3771         1.7749        1.4837
                             LAR       =          92.31     30.14    31.97    170.47   22.61    13.19    52.23     90.43         89.09




                                                                     76
CHAPTER 3                                                                                                             BS HUANG


Table 3.5:   Example of LAR calculation using EAR model (for HK females). EAR is the excess absolute risk, defined as the rate
             of cancer in an exposed population minus the rate of cancer in an unexposed population. Here it means the excess
             cancer incidence caused by 100 mSv radiation in a specific year. LAR is acquired by summing the EAR for each
             attained age (a).


              Keys:

              e=exposed age (years), a=attained age (years, from e+3 to 99), S(a)=probability of surviving until age a, S(a)/S(e)
              =the probability of surviving to age a on condition that he/she survives to age e, risk=EAR.




                                                              77
CHAPTER 3                                                                    BS HUANG


                                           age at exposure(years)
cancer site   0           10       20        30      40    50     60        70      80
male
stomach       848.3       564.4    371.9    241.2   228.0   203.4   165.5   115.6   62.4
colon         554.0       368.6    242.9    157.5   148.9   132.8   108.1   75.5    40.8
liver         447.2       297.9    197.7    130.6   127.7   119.9   104.4   79.2    47.1
lung          573.5       382.1    253.9    168.7   167.5   161.9   147.5   119.1   76.5
prostate      19.0        12.7     8.3      5.4     5.1     4.6     3.7     2.6     1.4
bladder       357.1       237.9    158.2    105.3   105.2   103.2   96.6    81.3    55.0
thyroid       -           -        -        -       -       -       -       -       -
other solid   1073.3      714.1    470.5    305.1   288.5   257.4   209.5   146.3   79.0

females
stomach       1067.1      709.5    468.0    304.5   290.5   264.6   222.8   162.7   92.3
colon         348.4       231.7    152.8    99.4    94.9    86.4    72.7    53.1    30.1
liver         272.3       181.2    120.2    79.4    77.9    74.1    66.0    51.8    32.0
lung          1193.2      794.2    527.4    350.0   347.5   338.1   312.8   258.4   170.5
uterus        261.3       173.7    114.6    74.6    71.1    64.8    54.6    39.9    22.6
ovary         152.4       101.4    66.9     43.5    41.5    37.8    31.8    23.2    13.2
bladder       324.9       216.2    143.6    95.5    95.2    93.7    88.6    75.7    52.2
other solid   1045.3      695.0    458.4    298.3   284.6   259.2   218.2   159.4   90.4
thyroid       -           -        -        -       -       -       -       -       -
breast        2026.9      1220.1   730.1    431.4   411.1   355.1   270.4   176.9   89.1

Table 3.6:    Summary of LAR results using EAR model (for the HK
              population). The data in the table is the numbers of cancer cases
              per 100,000 persons exposed to a radiation dose of 100mGy.


                  Keys:
                  LAR= lifetime attributable risk.




                                             78
CHAPTER 3                                                                BS HUANG


                                    age at exposure(years)
cancer site   0        10      20     30      40     50    60          70      80
  Males
stomach       139.1    99.9    71.5    50.6    48.7    44.3    36.0    25.0    11.5
colon         383.6    275.3   196.6   139.6   135.0   124.4   103.1   68.0    29.2
liver         228.6    163.2   116.2   82.1    76.8    65.5    48.3    28.5    12.8
lung          405.2    278.8   191.4   131.4   129.9   124.6   110.2   82.2    45.6
prostate      39.2     28.1    20.2    14.4    14.2    13.8    12.0    8.3     4.1
bladder       176.5    126.8   91.0    65.3    64.3    61.8    54.6    41.1    20.9
other solid   1609.4   532.8   340.5   222.7   195.7   159.8   116.1   67.6    27.9
thyroid       124.9    54.2    23.2    9.4     8.2     6.7     4.0     2.5     0.1

  Females
stomach       203.2    145.8   104.4   74.0    69.2    58.7    45.7    25.0    8.3
colon         164.7    116.7   83.1    59.0    53.8    48.3    38.4    22.0    8.2
liver         243.4    172.6   122.1   86.7    76.1    59.6    41.4    18.9    5.5
lung          831.3    567.1   389.1   267.0   258.4   244.0   220.3   166.9   95.4
breast        1351.2   813.4   486.7   287.6   274.1   236.7   180.2   118.0   59.4
uterus        6.3      4.5     3.2     2.3     2.1     1.6     1.0     0.4     0.2
ovary         56.6     40.4    28.1    19.5    16.1    11.8    7.8     4.5     1.5
bladder       203.7    141.4   101.3   72.8    72.0    70.7    65.2    52.8    32.7
other solid   1869.4   822.4   546.6   369.2   312.0   246.3   196.3   148.0   78.5
thyroid       735.5    318.9   132.9   54.0    43.6    32.5    21.5    11.9    1.9

Table 3.7:    Table of LAR of cancer incidence for the HK population (weighted
              results of EAR and ERR results ). Numbers of cases per 100,000
              persons exposed to a radiation dose of 100mSv are shown in this
              table.


               Keys:
               LAR= lifetime attributable risk.




                                       79
CHAPTER 3                                                                   BS HUANG


                                       age at exposure(years)
cancer site   0           10      20     30      40     50    60          70      80
  Males
stomach       66.1        56.6    34.8    24.4    23.5    21.8    17.4    12.2    6.1
colon         309.1       262.2   159.2   115.0   112.2   104.0   86.5    59.8    27.6
liver         70.8        58.0    34.8    25.5    24.4    22.0    16.2    9.3     3.5
lung          307.7       255.8   146.0   102.9   101.9   99.0    87.2    63.7    33.3
prostate      89.3        76.8    46.1    33.6    33.6    31.7    25.0    13.4    4.8
bladder       217.4       184.1   112.3   82.2    82.2    79.0    68.6    48.9    23.9
other solid   1111.8      665.3   308.9   196.0   170.3   138.6   97.0    56.4    22.8
thyroid       147.2       97.3    26.9    11.5    3.8     1.3     0.4     0.1     0.0

  Females
stomach       89.9        75.7    46.3    32.0    31.2    28.5    24.0    16.9    9.8
colon         206.8       175.8   107.2   77.1    74.3    68.6    58.3    42.3    21.6
liver         31.1        25.5    15.5    11.1    11.1    10.0    7.8     5.6     2.2
lung          777.0       644.5   366.8   256.5   254.4   243.8   213.1   155.8   81.6
breast        1089.0      850.0   399.0   235.3   131.1   65.1    28.8    11.2    3.7
uterus        46.5        39.1    24.2    16.7    14.9    12.1    8.4     4.7     1.9
ovary         82.2        68.7    39.5    26.9    24.5    19.8    14.2    8.7     4.0
bladder       218.4       185.4   112.3   81.4    80.3    76.2    65.9    48.4    24.7
other solid   1312.2      704.6   316.5   202.9   177.4   145.0   106.8   66.6    29.4
thyroid       855.9       565.7   152.6   55.4    18.9    5.4     1.4     0.4     0.0

Table 3.8:    Updated Table of LAR of cancer incidence for the US population.
              Numbers of cases per 100,000 persons exposed to a radiation dose
              of 100mSv are shown in this table.


                  Keys:
               LAR= lifetime attributable risk.




                                          80
CHAPTER 4                                                                   BS HUANG




CHAPTER 4                  DOSES AND CANCER RISKS OF

PEDIATRIC CARDIAC CT SCAN


4.1 Introduction

As discussed in Chapter 1, it is necessary to evaluate the radiation doses from the

cardiac examinations performed on children and the biological effects caused by

radiation. Although many studies have been done for cardiac CT scan using other

MDCT scanners, to the author’s knowledge, there have been no reports about dose to

children from cardiac CT scan performed on 64-slice CT scanners, nor cancer risk

estimation induced by these doses. Hence, in this part of our study, we aimed to (1)

measure the radiation dose from cardiac CT scan in children using 64-slice MDCT

with retrospective electrocardiograph (ECG)-gating and ECG-modulated tube current

technology, and calculate the effective dose; (2) evaluate the LAR of cancer

incidence for pediatric patients (about 5-year-old) in Hong Kong associated with the

radiation, and compare this with US patients. Special attention was paid to the

relationship between radiation dose (and then cancer risk) and heart rate. Four

different heart rates (40, 60, 70, 90 beats per minute, bpm) were studied and

compared. TLD-100H and the pediatric phantom introduced in Chapter 2 were used

for radiation dose measurement. The principles introduced in Chapter 2 were applied

for the estimation of radiation induced cancer risk.




                                          81
CHAPTER 4                                                                    BS HUANG



4.2 Literature review

4.2.1 Radiation exposure in cardiac pediatric CT scan

Applications of cardiac CT in children have been limited by respiratory motion,

small size of pediatric patients, and long sedation periods in the past. The benefits of

advancements in CT technology (refer to Chapter 1) have solved some of the

problems and have led to the increasing number of examinations in children (Cohen

et al, 2000; Frush et al, 2005; Siegel, 2005). Despite these advantages, successful

pediatric cardiac CT scanning in children mandates that careful attention be paid to

the balance between image quality and radiation dose (Slovis, 2002). Cardiac CT

scans have been well known to result in higher radiation dose compared to other

imaging modalities which involve ionizing radiation, as discussed in Chapter 1.


4.2.2 Cancer risk associated with cardiac pediatric CT scan

Cancer risks from pediatric CT scans have been widely reported, but few are for

cardiac CT. Brenner estimated lifetime cancer mortality risks attributable to the

radiation exposure from a CT in a 1-year-old are 0.18% for abdominal CT and 0.07%

for head CT (Brenner et al, 2001). Chodick et al reported that a lifetime mortality

risk of 0.29% would be associated with annual pediatric CT scanning from 1999 to

2003 in Isarel, which was thought to be small but not negligible (Chodick et al,

2007).


The Society for Pediatric Radiology organized a multidisciplinary conference to

discuss dose issues in pediatric CT in August 2001 (Slovis, 2003) and emphasized

the radiation dose requirement of “as low as reasonably achievable” (ALARA)


                                           82
CHAPTER 4                                                                  BS HUANG


concept in cardiac CT scan and that cancer risks associated with the radiation

exposure should be investigated. However, few studies have been done for pediatric

cardiac scan on the 64-slice MDCT to date.


4.3 Materials and methods

The 64-slice MDCT (LightSpeed VCT system, General Electric Healthcare Systems,

Milwaukee, WI) in our unit was studied. It is also used to study the doses of CT scan

in other parts of our research. To measure the organ doses, direct measurement of

radiation dose using TLD-100H and the anthropomorphic 5-year-old pediatric

phantom described in Chapter 2 were used. From the measured organ doses, the risks

for common cancer induction were calculated by using the age dependent LARs of

cancer incidence from the table set up for the HK population described in Chapter 3.

The cancer risks for 5-year-old US children were also estimated and a comparison

between the HK and the US population was made.


4.3.1 Organ-specific CT dose measurement

A total number of 58 TLDs were distributed in various organs of interest in the

phantom (refer to Chapter 2) and 8 TLDs were used as controls (Table 4.1). To

measure radiation dose to skin, the TLDs were placed inside small black plastic bags

which were then attached to the skin surface in the directly scanned region (chest and

back). The readings of a specific organ were averaged to calculate the absorbed dose

of this organ.



Protocols of retrospectively ECG-gated cardiac CT scan for pediatric patients with

different heart rates, 40 bpm, 60 bpm, 70 bpm and 90 bpm were studied with a pitch


                                         83
CHAPTER 4                                                                    BS HUANG


of 0.16, 0.22, 0.24 and 0.24 respectively. The other parameters were the same: 64×

0.625 mm configuration; 0.35 second gantry cycle time; 100 kV tube potential;

electrocardiograph (ECG)-modulated 200 mA ~ 280 mA (280 mA from 65% to 80%

of the cardiac cycle, 200 mA for the rest). The ECG-modulation technique only

provides full tube current in the diastolic period, therefore decreases the dose to

patients, as introduced in Chapter 1. The images were reconstructed using the data

acquired from 65% to 85% of heart beat cycle, according to retrospectively

ECG-gated image reconstruction technology.



The total effective dose from pediatric cardiac CT scan was calculated by using the

methods described in Chapter 2.


4.3.2 Cancer risk estimation

Cancer risk estimation was performed by applying the methods introduced in

Chapter 2, according to the organ doses and the cancer risk estimating table

established for the HK population and the updated table for the US population in

Chapter 3. We repeated the analysis of cancer risk for both boys and girls of 5 years

of age



To evaluate the impact of the associated cancer risk, the proportion of LAR caused

by radiation in total lifetime cancer incidence was calculated.


4.4 Results

4.4.1 Radiation doses

Organ doses and effective doses measured are summarized in Table 4.2. The

                                           84
CHAPTER 4                                                                      BS HUANG


effective doses of heart rates of 40 bpm, 60 bpm, 70 bpm, and 90 bpm for 5-year-old

girls were 16.45 mSv, 12.18 mSv, 11.98 mSv and 11.82 mSv respectively; and for

5-year-old boys were 16.46 mSv, 12.17 mSv, 11.97 mSv and 11.80 mSv, respectively.

A high radiation burden was found to the heart, lung, breast, thymus, and esophagus,

which were organs directly exposed to radiation. The other organs, which were only

partially scanned directly or not at all, contributed to less than 10 percent of the total

effective dose.


4.4.2 LAR of Cancer Incidence

LAR of cancer incidence of the HK and the US population induced by the radiation

dose presented above is illustrated in Table 4.3 and Figure 4.1. The excessive cancer

incidence (LAR) for the four heart rates (40 bpm, 60 bpm, 70bpm and 90 bpm) for

5-year-old Hong Kong boys associated with the cardiac CT scan was 0.33% (1 in

303), 0.24% (1 in 417), 0.22% (1 in 455) and 0.22% (1 in 455); for 5-year-old Hong

Kong girls was 0.85% (1 in 118), 0.62% (1 in 161), 0.63% (1 in 159) and 0.61% (1 in

164) respectively. The LARs from the cardiac CT scan with the four heart rates for

5-year-old US boys were 0.20%, 0.15%, 0.14% and 0.14% respectively; for

5-year-old US girls were 0.60%, 0.44%, 0.45% and 0.43% respectively.



The risks were 150%~230% higher in girls than in boys, and 41%~63% higher in US

compared to HK children. Girls’ breast, lung, and bone marrow were the three organs

with highest risk of cancer incidence, and for boys the three organs were lung, bone

marrow and liver.



The baseline lifetime cancer incidence for 5-year-old boys in Hong Kong was


                                            85
CHAPTER 4                                                                 BS HUANG


32.87%, and for 5-year-old girls was 23.32%, as shown in Table 2.9 in Chapter 2.

Hence, LARs from radiation dose for the four protocols contributed to 0.99%, 0.73%,

0.67% and 0.67% of the overall cancer incidence respectively for Hong Kong boys;

and 3.51%, 2.61%, 2.64% and 2.56% respectively for Hong Kong girls. The baseline

lifetime cancer incidence for 5-year-old boys in US is 43.89%, and for 5-year-old

girls is 37.35%. Hence, LARs from radiation dose for the four protocols contributed

to 0.46%, 0.34%, 0.31% and 0.31% of the overall cancer incidence respectively for

US boys; and 1.57%, 1.16%, 1.18% and 1.15% respectively for US girls. These

results are shown in Table 4.4 and Figure 4.2.


4.5 Discussion

Despite the inherent limitations in methodology (refer to Chapter 2), our dose results

provide an opportunity to compare the doses with pediatric cardiac CT scans

employing different scanners, protocols and technologies in the literature.

Hollingsworth et al estimated the radiation dose of pediatric retrospectively

ECG-gated coronary CTA to be around 26 mSv using a high dose protocol (120kV,

330mA, pitch of 0.275, 16×0.625mm, 0.5s rotation time, small field of view) on a

16-slice CT scanner (LightSpeed, GE Healthcare), according to tissue weighting

factors in ICRP publication 60 (Hollingsworth et al, 2007). We recalculated the

effective dose to be 34 mSv according to tissue weighting factors in ICRP

publication 103. The recalculated effective dose was much higher using ICRP 103

than ICRP 60, because the tissue weighting factors are markedly different between

the two, and the highest factors in ICRP 103 are for lung, stomach, breast and bone

marrow, all of which are partly or wholly exposed to radiation in cardiac CT. For the

other protocols in the work of Hollingsworth, the recalculated effective doses were


                                         86
CHAPTER 4                                                                BS HUANG


22 mSv (120 kV, 220 mA), 10 mSv (120 kV, 110 mA) and 14 mSv (80kV, 385mA).

These dose results were higher than our results (except for 110mA protocol), due to a

higher tube potential (or higher tube current, for 80kV protocol) and longer rotation

time (0.5s VS 0.35s) and without ECG-controlled tube current modulation. Herzog C

et al reported a substantial radiation dose reduction of pediatric cardiovascular

64-slice MDCT angiography using automatic anatomic tube current modulation.

Scan protocol was; 64×0.6 mm configuration; 0.33 second rotation time; pitch of 1.5;

80 kV for children with body weight less then 15 kg, 100 or120 kV for children with

a weight above 15 kg; base low tube current of 72 mAs (Herzog et al, 2008). The

effective doses measured were 1.0 mSv, 1.9 mSv, and 4.4 mSv for 80 kV, 100kV, and

120 kV respectively according to ICRP publication 60 (Herzog et al, 2008), which

were recalculated according to ICRP publication 103 to be about 1.3 mSv, 2.5mSv,

and 5.8 mSv respectively. The low doses are due to the high pitch of 1.5, base low

tube current of 72 mAs, and the use of automatic anatomic tube current modulation.

Horiguchi et al reported that the prospective ECG-gated technology reduced the

effective dose to about 3.0 mSv from 13 mSv imparted by retrospectively ECG-gated

CTA with the same tube current and potential (Horiguchi et al, 2008). Similarly, low

mean effective doses of 2.1±0.6 mSv were published by Husmann et al from

prospective ECG-gated CTA in adult patients (Husmann et al, 2008). This technique

principally acquires data only when it is necessary for image reconstruction, and

minimizes the overlapping data acquisition. However, the application of prospective

ECG-gated CTA is limited on models of advanced CT scanners currently due to its

high requirement of hardware and software (Schoenhagen, 2008).



Thus, published results suggest that lower radiation dose can be achieved by


                                         87
CHAPTER 4                                                                 BS HUANG


applying low-dose protocol parameters and still maintaining diagnostic quality

images. Dose reduction can also be realized by routinely applying prospective

ECG-gating technology and automatic anatomic tube current modulation technology

in the future.



The lifetime cancer risks calculated for cardiac CT scan were higher compared to the

risks associated with other CT examinations that have been reported in the literature,

including low-dose chest CT (0.06%), abdominal CT (0.18%) and head CT (0.07%)

(Brenner et al, 2001; Huda, 2007). Our estimated cancer risks were also higher

compared to the cancer risks from adult cardiac CT examinations. For example,

Einstein et al (Einstein et al, 2007) reported a LAR of 0.46%, 0.23%, 0.15% for 20,

40, 60 years old females respectively, and less than 0.1% for adult males of all age

groups associated with 64-slice MDCT coronary angiography (120 kV, 170 mAs,

0.33 seconds rotation time, pitch of 0.2, electrocardiographically controlled tube

current modulation). The higher risks in our study was mainly due to the more severe

radiation-induced detrimental effect in children than in adults (ICRP, 1991; ICRP,

2007; National Research Council of the National Academies, 2006; UNSCEAR,

2000a) as the effective doses in our results and the work of Einstein et al are mainly

the same (ours are from 11.8 mSv to 16.5 mSv, while the recalculated effective dose

in Einstein’s research is a little higher, about 16.8 mSv).



The risks estimated for HK children are higher than US children, as shown in Figure

4.1. This is mainly due to the differences in life table and cancer statistics data

between the two populations. Firstly, a higher cancer incidence is observed for the

HK population in the organs which are wholly or partly exposed to X-ray radiation in


                                           88
CHAPTER 4                                                                  BS HUANG


cardiac CT scan (Table 4.5). For example, the HK population suffers a higher cancer

incidence in the lung, colon, liver, and stomach compared to the US population,

although HK females have lower breast cancer incidence than US females. Secondly,

the life expectancy of the HK population is longer than Americans, and therefore

have a longer period to develop cancer after exposure to radiation (HK Census and

Statistics Department, 2006; US Disaster Center, 2006).


Our results showed that the proportion of LAR in total lifetime cancer incidence

from the radiation exposure was more than 0.67% and 2.56% for HK boys and girls

respectively; and more than 0.31% and 1.15% for US boys and girls respectively,

which less than half of that for HK children. Hence the risks estimated are considered

to be high and should not be neglected. This is due to a higher LAR of cancer

incidence but a lower baseline lifetime risk (Table 2.9, 2.10) in HK children.



The doses and cancer risks for the four heart rates were different due to the

difference in pitches. Pitch varies with heart rate while other parameters of the

cardiac protocol such as tube current and potential have no relation with heart rate.

Pitch decreases when heart rate increases from 40 bpm to 70 bpm but remains the

same between 70 bpm and 90 bpm. Pitch settings have great impact on the radiation

exposure to patients (Goldman, 2008a). Increasing pitch would reduce the radiation

dose and consequently lower the cancer risk to patients. Hence, for the

ECG-modulated cardiac CT scans we studied, the results in Table 4.2, 4.3 and Figure

4.1 showed that the radiation dose and cancer risk would mainly decrease when the

heart rate increases, although the reduction in dose is much less when heart beat is

above 60 bpm. However, high heart rates would affect image quality as the temporal

resolution of CT scanners are limited even for 64-slice MDCT which is already

                                          89
CHAPTER 4                                                                  BS HUANG


recognized as providing relatively high temporal resolution (Desjardins et al, 2004).



It was reported that the awareness of radiation dose imparted by diagnostic imaging

to children was generally low among pediatricians with widespread underestimation

of radiation exposure and the related risks (Thomas et al, 2006). This underscores the

importance of improving the knowledge and understanding of radiation dose and its

effects in children. Our study provides valuable information on this subject and will

contribute to awareness among pediatricians. The studies are essential to provide

accurate information to parents when obtaining informed consent for such a

procedure.


4.6 Summary and conclusion

In our research the radiation dose and cancer risks for 5-year-old children associated

with the cardiac pediatric CT examinations for a series of heart rates were

investigated. Our results showed that the doses and cancer risks for pediatric patients

were high compared to other CT application and baseline cancer incidence, and this

was higher in HK children compared to US children. Dose and cancer risk increased

when heart rate decreased. It is suggested that protocol parameters should be

optimized and radiation dose to children should be lowered to the ALARA level.

Possible dose reduction techniques include applying low-dose protocols, prospective

ECG-gating technology and automatic anatomic tube current modulation technology.




                                          90
CHAPTER 4                                                                                    BS HUANG



Figures:


                                                                                   Hong Kong girls
                                  1.0%
        LAR of Cancer Incidence                                                    American girls
                                                                                   Hong Kong boys
                                  0.8%
                                                                                   American boys

                                  0.6%

                                  0.4%


                                  0.2%

                                  0.0%
                                                 40          60             70        90
                                                              Heart Rate (bpm)



Figure 4.1:                              LAR of cancer incidence induced by pediatric cardiac CT scan.


                                         Keys:
                                         LAR= lifetime attributable risk.




                                                                  91
CHAPTER 4                                                                                        BS HUANG



                                                                                    Hong Kong girls
                                      4.0%
                                                                                    American girls




         Proportion of LAR in Total
                                                                                    Hong Kong boys
                                      3.0%                                          American boys


                                      2.0%


                                      1.0%


                                      0.0%
                                                 40           60             70          90
                                                               Heart rate (bpm)


Figure 4.2:                            The proportion of LAR in total cancer incidence. The total cancer
                                       incidence is acquired by adding the LAR to baseline lifetime
                                       cancer incidence (refer to section 2.6 in Chapter 2)


                                        Keys:
                                        LAR= lifetime attributable risk.




                                                                   92
CHAPTER 4                                                      BS HUANG



Tables:


                   organ            number of TLDs
                   background       8
                   lens             2
                   skin             8
                   brain            2
                   salivary gland   1
                   thyroid          1
                   thymus           1
                   esophagus        3
                   lung             11
                   bone surface     3
                   heart            3
                   muscle           3
                   bone marrow      3
                   stomach          3
                   liver            3
                   colon            2
                   pancreas         1
                   spleen           1
                   small bowel      2
                   adrenal          1
                   kidney           1
                   bladder          1
                   ovary            1
                   uterus           1
                   testicle         2

Table 4.1:   Thermoluminescent dosimeters (TLD) distribution in pediatric
             phantom.




                                    93
CHAPTER 4                                                           BS HUANG



             organ           WT     40bpm       60bpm   70bpm     90bpm
     gonad
     ovary (girls)          0.08    0.2         0.2     0.2      0.2
     testicle (boys)                0.1         0.1     0.1      0.1
     colon                   0.12   4.0         3.1     2.8      3.0
     stomach                 0.12   24.3        17.8    18.7     19.1
     breast                  0.12   28.5        20.9    25.1     23.3
     lung                    0.12   27.8        19.9    17.7     18.1
     bone marrow             0.12   21.5        17.4    14.3     13.8
     thyroid                 0.04   4.0         2.8     2.6      2.6
     liver                   0.04   20.8        14.3    13.5     14.3
     esophagus               0.04   23.6        17.4    15.8     15.8
     bladder                 0.04   0.2         0.2     0.2      0.2
     skin                    0.01   28.4        20.9    18.8     18.6
     bone surface            0.01   33.7        26.3    20.8     20.2
     brain                   0.01   0.2         0.2     0.1      0.2
     kidney                 0.013   2.7         2.0     2.7      2.7
     spleen                 0.013   5.2         5.2     5.2      5.2
     adrenal                0.013   3.0         2.2     3.0      3.0
     prostate               0.013   0.2         0.2     0.2      0.2
     pancreas               0.013   6.9         3.6     6.9      6.9
     small bowel            0.013   2.5         1.5     1.5      1.5
     thymus                 0.013   12.6        9.8     9.5      9.5
     muscle                 0.013   24.3        21.2    18.1     17.9
     heart                  0.013   37.7        21.8    24.8     21.8
     salivary gland         0.013   0.9         0.7     0.7      0.7
     effective dose (mSv)
     girls                          16.57       12.26   12.06    11.90
     boys                           16.56       12.25   12.05    11.88

Table 4.2:   CT doses for boys and girls (in mSv) from pediatric cardiac CT
             scan. Four heart rates were programmed into the ECG generator.
             Effective doses were calculated using the tissue weighting factors
             (WT) recommended in ICRP publication 103.


               Keys:

               bpm = beats per minute.




                                           94
CHAPTER 4                                                       BS HUANG



                           boys’ LAR                      girls’ LAR
 heart rate/bpm   HK             US              HK              US
 40               0.328%         0.203%          0.847%          0.596%
 60               0.243%         0.150%          0.624%          0.438%
 70               0.221%         0.136%          0.632%          0.448%
 90               0.222%         0.136%          0.613%          0.434%

Table 4.3:   LAR of cancer incidence induced by pediatric cardiac CT scan.
             Keys:


              LAR= lifetime attributable risk.




                                     95
CHAPTER 4                                                             BS HUANG



             heart rate                 LAR proportion
             (bpm)        HK boys     HK girls US boys     US girls
             40           0.99%       3.51%    0.46%       1.57%
             60           0.73%       2.61%    0.34%       1.16%
             70           0.67%       2.64%    0.31%       1.18%
             90           0.67%       2.56%    0.31%       1.15%

Table 4.4:     The proportion of LAR induced by pediatric cardiac CT scan in
               total cancer incidence. The total cancer incidence is acquired by
               adding the LAR to baseline lifetime cancer incidence (refer to
               section 2.6 in Chapter 2)


                  Keys:
                  LAR= lifetime attributable risk.




                                           96
CHAPTER 4                                                             BS HUANG



               HK                                        US
                 relative                                  relative
rank site                        WT      rank site                          WT
                 frequency                                 frequency
1    lung        18.20%          0.12    1    prostate     15.40%           0.013
2    colorectum 16.30%           0.12    2    breast       14.66%           0.12
3    breast      10.20%          0.12    3    lung         13.67%           0.12
4    liver       7.70%           0.04    4    colorectum   10.83%           0.12
5    stomach     4.50%           0.12    5    bladder      4.54%            0.04

Table 4.5:   Top five cancer sites of US and the HK populations. HK data is
             from Hong Kong government (Hong Kong Cancer Registry, 2005),
             and US data is from National Cancer Institute (US National
             Cancer Institute, 2006a). Relative frequency is defined as the cases
             for this site divided by all the cancer cases. WT is the
             corresponding tissue weighting factor recommended in ICRP
             publication 103. As shown in this table, the leading cancer sites in
             Hong Kong have higher WT, i.e. more sensitive to radiation dose,
             than the leading cancer sites in US.




                                        97
CHAPTER 5                                                                 BS HUANG




CHAPTER 5 DOSES AND CANCER RISKS OF ADULT

CORONARY CT ANGIOGRAPHY


5.1 Introduction

The objectives for this part of our research on adult CT angiography (CTA) were: (1)

to estimate the radiation dose from coronary CTA in adults using 64-slice MDCT; (2)

to evaluate and compare the LAR of cancer incidence for the adult patients in Hong

Kong and United States associated with the radiation. Radiation dose was simulated

using the spreadsheet ImPACT, as shown in Chapter 2. Radiation induced cancer

risks were estimated according to the principles in BEIR VII report, as introduced in

Chapter 2. Three protocols with pitches, 0.2, 0.22, and 0.24 were studied and

compared.




                                         98
CHAPTER 5                                                                  BS HUANG



5.2 Literature review

Radiation dose has become an important issue for cardiac MDCT (coronary CTA)

since the advent of angiography using MDCT (Leschka et al, 2005; Mollet et al,

2005). The reported doses from coronary CTA on adults were much higher than

other types of CT scan, as shown in Table 1.2. Effective dose was estimated to be 8.1

mSv (male) and 10.9 mSv (female) for coronary CTA performed on a 16-slice

MDCT (120 kV, 400 mAs, 12 x 0.75 mm) (Trabold et al, 2003). It is also reported

that the average effective dose for males and females undergoing coronary CTA on a

16-slice MDCT scanner was 12.05 mSv (120 kVp, 550 mAs, 0.28 pitch) (Hohl et al,

2006). For 64-slice MDCT, the lowest average effective dose of coronary CTA was

reported to be 9.5±3.4 mSv per patient (range 7.1 mSv-17.7 mSv) (Francone et al,

2007), and up to 21 mSv for females and 15 mSv for males on the same scanner

(Einstein et al, 2007). On dual-source CT, the mean effective dose that resulted from

coronary CTA was reported to be from 7.8 mSv to 8.8 mSv (Stolzmann et al, 2008),

which was 50% lower than the effective doses estimated by Einstein et al (Einstein et

al, 2007). The organ doses from coronary CTA, especially for the organs exposed

directly to X-ray are much higher, up to 114 mSv for esophagus, 80 mSv for breast,

and 91 mSv for lung (Einstein et al, 2007).



The cancer risks associated with CTA in adults has been reported in the literature, as

described in section 4.5, Chapter 4. A higher cancer risk is expected in coronary CTA

compared to other types of CT scan (Table 5.1). To date, information on radiation

induced cancer risk is scarce with only two papers addressing the subject in coronary

CTA (Einstein et al, 2007; Einstein et al, 2008). Also, both the scanner models


                                          99
CHAPTER 5                                                                   BS HUANG


studied in these two publications are different from the scanner used in our institution.

Therefore, in this part of our study, we aimed to investigate the dose and then the

cancer risk to adults associated with coronary CTA on the GE 64-Slice MDCT in our

institution, which would be valuable information for both the patients and clinicians

in our institution.


5.3 Methodology

5.3.1 Organ-specific CT dose estimation

To estimate the radiation dose of CTA, the spreadsheet ImPACT was applied, as

introduced in Chapter 2. The 64-slice LightSpeed VCT was also studied. The helical

protocols of angiography were: 120 kV; 0.35 second rotation time; cardiac large filter;

0.625 mm slice thickness; 350 mA~700 mA ECG-modulated tube current (full

current between 40%~80% of heart beat period); pitch of 0.2 or 0.22 or 0.24. The

effective doses were calculated according to the organ doses and weighting factors

recommended in ICRP publication 103, as introduced in Chapter 2.


5.3.2 Calculation of lifetime cancer incidence

Cancer risk estimation was performed by applying the method introduced in Chapter

2, according to the dose results and the cancer risk estimating table established

(updated) for the HK (US) population in Chpater 3. We repeated the analysis of

cancer risk for Hong Kong and US adult patients with a range of ages between 20

and 80 years old.



To study how severe the associated cancer risk was, the proportion of LAR caused by

the radiation in total lifetime cancer incidence was calculated, according to method

                                          100
CHAPTER 5                                                                BS HUANG


introduced in Chapter 2.


5.4 Results

5.4.1 Radiation doses

Organ doses and effective doses from CTA using VCT are summarized in Table 5.2.

According to the tissue weighting factors of ICRP publication 103, for females the

effective doses of the protocols with pitch of 0.2, 0.22, 0.24 were 23.32 mSv, 21.05

mSv, and 20.76 mSv respectively; and for males, were 27.33 mSv, 24.80 mSv, 22.85

mSv respectively, higher than that of females. The organ doses and effective doses

decrease when the pitch increases. A high radiation burden was found in heart, lung,

breast, thymus, and esophagus, which were the organs directly exposed to the

radiation. Doses imparted to these directly-exposed organs ranged from 59 mSv to 73

mSv. The organs which were not scanned directly contributed to less than 6% of the

overall effective dose, such as gonad, colon, and bladder.


5.4.2 Cancer risks estimated for the HK and US population

LAR of cancer incidence induced by the radiation dose is illustrated in Table 5.3 and

Figure 5.1, and this decreased when pitch increased. LAR in 20-year-old males in

HK undergoing a single CTA was 0.226% (1 in 442), 0.205% (1 in 488) to 0.189% (1

in 529) for pitch 0.2, 0.22 and 0.24 respectively. For 20-year-old females in HK it

was 0.690% (1 in 145), 0.623% (1 in 161) and 0.585% (1 in 171) respectively. The

risks were high in younger ages and decreased with age for both genders. For

example, for pitch 0.2, LAR of males decreased from 0.226% at 20-years of age to

0.041% at 80-years of age. LARs for women were higher than men across all ages.

This was mainly because girls’ breasts receive a high dose and carries a high risk of

                                         101
CHAPTER 5                                                                   BS HUANG


developing cancer.



The risks for the HK population were higher than for the US population for both

genders and all ages, as shown in Table 5.3 and Figure 5.1. The LAR in 20-year-old

males in US undergoing a single CTA was from 0.169% (1 in 592), 0.153% (1 in 654)

to 0.141% (1 in 709) for pitch 0.2, 0.22 and 0.24 respectively. For 20-year-old

females in US it was 0.550 % (1 in 182), 0.490% (1 in 204) and 0.460% (1 in 217)

respectively.


5.4.3 Comparison with baseline lifetime cancer risk

The baseline lifetime cancer incidence for the HK and the US population is listed in

Table 2.9 and 2.10. The proportion of LAR in total lifetime cancer risk was

calculated. The results are shown in Figure 5.2. For the HK population the proportion

was more than 0.35% and 1.5% for males and females respectively; for the US

population it was more than 0.15% and 0.4% for males and females respectively.


5.5 Discussion

There are inherent limitations to our results of doses and cancer risks, mainly due to

the limitations in methodology (refer to Chapter 2). Our dose results are comparable

to some published results of CTA performed on 64-slice MDCT scanners. Besides

the effective doses of 21 mSv for females and 15 mSv for males estimated by

Einstein et al (refer to section 5.2), the average effective dose with retrospective CTA

was estimated to be 20.0±3.5 mSv (Hirai et al, 2008) by a recent study in Japan. The

effective dose in our study recalculated according to ICRP publication 60 used by

Einstein and Hirai is up to 21 mSv for males and 18 mSv for females (for pitch 0.2).


                                          102
CHAPTER 5                                                                    BS HUANG


This is in good accordance with the results by Hirai et al and Einstein et al. Our

results are much higher than the effective doses reported for prospective ECG-gated

CTA, which was found to be only about 20%~25% of the doses of retrospective

ECG-gated CTA (Hirai et al, 2008; Horiguchi et al, 2008; Shuman et al, 2008).

However, prospective ECG-gated technology is available only in advanced CT

scanners. Our results are also at least 100% more than the reported effective dose of

CTA on dual-source CT, which was found to be about 7.8 to 8.8 mSv (Stolzmann et

al, 2008). However further studies are needed to evaluate and compare image quality

and diagnostic accuracy of dual-source CT and other 64-slice MDCT scanners.



Cancer risks for the US population estimated in our study were also similar with the

work by Einstein et al in which the LAR associated with coronary CTA on 64-slice

MDCT was reported to be 0.70% and 0.15% for 20-year-old US women and men

respectively (Einstein et al, 2007). This is due to the similar dose results in our study

and the work of Einstein et al. As expected, the risks estimated in our study were

higher than the risks from other types of CT scan (Table 5.1).



Both the doses and risks change markedly with pitch. As seen in Table 5.2, the

effective dose of males for pitch 0.22 is 9.3% lower than for pitch 0.2, and 8.5%

higher than for pitch 0.24. The LAR for pitch 0.2 is 10.8% higher than for pitch 0.22,

and 17.9% higher than pitch 0.24 for females. However the selection of pitch should

be balanced with compromise in image quality (Nakanishi et al, 2005).



Cancer risks to the HK population were higher than the US population, consistent

with the results in Chapter 4. The reasons were the same as provided in Chapter 4:


                                          103
CHAPTER 5                                                                 BS HUANG


the longer life expectancy of the HK population and different baseline cancer

incidence in the organs. As Americans have a higher baseline lifetime cancer

incidence than Hong Kong-ers (Table 2.9 and 2.10) but a lower LAR associated from

the same radiation dose, the proportion of LAR in total cancer incidence was lower

in the US population than in the HK population. Hence it was concluded than one

single coronary CTA was more harmful to the HK population than to the US

population.


5.6 Summary and conclusion

In this study, the radiation doses from retrospectively ECG-gated coronary CT scans

were simulated, and the cancer incidence risks associated with the dose were

estimated. These risk estimates provide valuable information for considering the

risk-benefit balance for coronary CTA.



It is concluded that 64-slice CTA is accompanied by high cancer risk. Such cancer

risk depends on the protocols used, such as pitch, tube current and potential, and the

scanner types. Cautions should be taken by doctors and patients when they make

such medical imaging.




                                         104
CHAPTER 5                                                                                                                           BS HUANG


Figures:
                                                                                                                HK   females




                                      LAR of Cancer Incidence
                                                                0.8%
                                                                                                                US   females
                                                                0.6%                                            HK   males
                                                                                                                US   males
                                                                0.4%

                                                                0.2%

                                                                0.0%
                                                                            20    30        40     50    60    70        80

                                                                                      Exposed Age (years)


                                                                                      (a)    pitch=0.2
                                      LAR of Cancer Incidence




                                                                0.8%                                            HK   females
                                                                                                                US   females
                                                                0.6%                                            HK   males
                                                                                                                US   males
                                                                0.4%

                                                                0.2%

                                                                0.0%
                                                                            20    30        40     50    60    70        80
                                                                                      Exposed Age (years)


                                                                                      (b)   pitch=0.22
           LAR of Cancer Incidence




                                     0.8%                                                                            HK       females
                                                                                                                     US       females
                                     0.6%                                                                            HK       males
                                                                                                                     US       males
                                     0.4%

                                     0.2%

                                     0.0%
                                                                       20        30         40     50     60        70         80

                                                                                  Exposed Age (years)


                                                                                      (c)   pitch=0.24

Figure 5.1:                                           LAR of cancer incidence induced by coronary CT angiography.


                                               Keys:
                                               LAR= lifetime attributable risk.


                                                                                             105
CHAPTER 5                                                                        BS HUANG



                                                                     HK female
                            3.0%                                     US female
                                                                     HK male
                                                                     US male
                            2.5%

                            2.0%




              percentage
                            1.5%

                            1.0%

                            0.5%

                            0.0%
                                    20    30      40      50    60   70
                                                  exposed age

                                            (a)   pitch=0.2

                                                                     HK female
                            3.0%                                     US female
                                                                     HK male
                                                                     US male
                            2.5%

                            2.0%
              percentage




                            1.5%

                            1.0%

                            0.5%

                            0.0%
                                    20    30      40      50    60   70
                                                  exposed age

                                           (b)    pitch=0.22

                                                                     HK female
                            3.0%                                     US female
                                                                     HK male
                                                                     US male
                            2.5%

                            2.0%
              percentage




                            1.5%

                            1.0%

                            0.5%

                            0.0%
                                   20     30      40      50    60   70
                                                  exposed age

                                           (c)    pitch=0.24

Figure 5.2:      The proportion of LAR in total cancer incidence. The total cancer
                 incidence is acquired by adding the LAR to baseline lifetime
                 cancer incidence (refer to section 2.6 in Chapter 2)


                           Keys:
                           LAR= lifetime attributable risk.

                                                   106
CHAPTER 5                                                               BS HUANG


Tables:


                            effective dose/mSv lifetime excess cancer incidence
study type
                           (average of genders)      (20 years old patients)
head CT                     8.7 (thyroid dose)         0.039% (thyroid)
chest CT                            5.4                     0.025%
coronary CT angiography            23~29                 0.15%~0.71%
pediatric abdominal CT
                                    11                         0.18%
(1 year old)
whole body CT                      12.1                  0.135% (mortality)

Table 5.1:   Typical cancer risks from various radiologic studies. Data were
             from: (1) head CT (Mazonakis et al, 2007); chest CT (Brenner et al,
             2004; Diederich et al, 2000; ICRP, 1991); CT angiography
             (Brenner et al, 2004; Einstein et al, 2007; Frush et al, 2003).




                                         107
CHAPTER 5                                                            BS HUANG




                                                dose(mSv)
                                pitch=0.2         pitch=0.22       pitch=0.24
 organ              WT      male     female    male     female   male female
 gonad              0.08    0.01     0.13      0.01     0.11     0.01 0.11
 colon              0.12    0.16     0.14      0.15     0.12     0.13 11
 stomach            0.12    11       9.6       10       8.7      9.4    8
 lung               0.12    72       61        65       55       60     51
 thyroid            0.04    2.2      1.9       2        1.7      1.9    1.6
 liver              0.04    19       16        17       14       15     13
 bladder            0.04    0.04     0.03      0.03     0.03     0.03 0.03
 breast             0.12    73       62        66       56       61     52
 uterus             0.013   0.13     0.11      0.12     0.1      0.11 0.09
 bone marrow        0.12    15       13        14       12       13     11
 bone surface       0.01    27       23        25       21       23     19
 skin               0.01    10       8.9       9.5      8.1      8.7    7.4
 brain              0.01    0.1      0.09      0.09     0.08     0.09 0.07
 thymus             0.013   71       61        65       55       59     50
 kidney             0.013   4.2      3.5       3.8      3.2      3.5    2.9
 pancreas           0.013   15       13        13       11       12     11
 spleen             0.013   13       11        11       9.8      11     8.9
 adrenal            0.013   23       20        21       18       19     17
 esophagus          0.04    71       61        65       55       59     50
 small intestines   0.013   0.67     0.57      0.61     0.52     0.56 0.48
 large intestines   0.013   0.94     0.8       0.85     0.72     0.78 0.66
 muscle             0.013   10       8.9       9.5      8.1      8.7    7.4
 heart              0.013   72       61        65       55       60     51
 effective dose        -    27.33 23.32        24.80 21.05       22.85 20.76

Table 5.2:    Doses from coronary CT angiography simulated using ImPACT.
              Effective doses were calculated using the tissue weighting factors
              (WT) recommended in ICRP publication 103.




                                      108
CHAPTER 5                                                          BS HUANG


                                 males’ LAR            females’ LAR
       exposed age (years)      HK         US         HK          US
               20             0.226%     0.169%     0.690%      0.550%
               30             0.154%     0.115%     0.443%      0.352%
               40             0.147%     0.110%     0.415%      0.280%
               50             0.135%     0.102%     0.365%      0.225%
               60             0.112%     0.084%     0.300%      0.175%
               70             0.079%     0.059%     0.213%      0.119%
               80             0.041%     0.029%     0.114%      0.060%
                                (a) pitch=0.2
                                  males’ LAR           females’ LAR
       exposed age (years)      HK          US        HK          US
               20            0.205%      0.153%    0.623%      0.496%
               30            0.140%      0.105%    0.400%      0.318%
               40            0.133%      0.100%    0.374%      0.253%
               50            0.122%      0.092%    0.329%      0.203%
               60            0.102%      0.077%    0.271%      0.158%
               70            0.072%      0.053%    0.192%      0.107%
               80            0.037%      0.026%    0.103%      0.054%
                                (b) pitch=0.22
                                  males’ LAR           females’ LAR
       exposed age (years)      HK          US        HK          US
               20             0.189%      0.141%    0.585%      0.460%
               30             0.128%      0.096%    0.377%      0.295%
               40             0.122%      0.092%    0.353%      0.234%
               50             0.112%      0.085%    0.310%      0.188%
               60             0.094%      0.070%    0.255%      0.146%
               70             0.066%      0.049%    0.180%      0.099%
               80             0.034%      0.024%    0.096%      0.050%
                                (a) pitch=0.24

Table 5.3:   LAR of cancer incidence induced by coronary CT angiography.
             Keys:


               LAR= lifetime attributable risk.




                                     109
CHAPTER 6                                                                   BS HUANG




CHAPTER 6 DOSES AND CANCER RISKS OF WHOLE

BODY PET-CT SCAN


6.1 Introduction

In this part of our study we aimed to (a) measure and estimate the radiation exposure

of patients undergoing whole-body 18F-FDG PET-CT examinations; (b) evaluate the

lifetime cancer risk to United States (US) and Hong Kong (HK) patients induced by

the radiation exposure. We studied the influence of different protocol parameters on

dose and cancer risk, impact of age and sex on cancer incidence, and the comparison

between the associated cancer risk and baseline incidence.




6.2 Literature review

Radiation exposure imparted to patients by 18F-FDG PET-CT examinations, which is

a combination of doses from CT scan and PET scan, is higher compared to dose from

stand alone CT or PET examinations. From the literature, this dose varied markedly
                                          18
depending on CT protocols and injected         F-FDG activities. According to the study

in Japan by Ghotbi et al, the total effective dose from the whole-body PET-CT

examinations was estimated to be 6.34 to 9.48 mSv for the average Japanese

individual at 60 kg body weight (Ghotbi et al, 2007) using screening CT protocols

and an average FDG activity of 370 MBq. Brix et al reported an effective dose from

whole-body PET-CT examinations to be 26.4 mSv using a diagnostic CT protocol

(140 kVp, 150mAs, pitch factor of 1.5) and also an average FDG activity of 370

MBq (Brix et al, 2005).
                                         110
CHAPTER 6                                                                     BS HUANG




The radiation exposure of some organs from PET-CT scan was specifically studied,

such as breast. Halac et al measured the doses to female breast using TLD placed on

the surface of the breasts and a mean dose of 14.42 ± 2.41 mSv was reported (Halac

et al, 2007). The radiation dose imparted to technicians in the PET-CT unit has also

been reported. Based on an annual number of less than 500 patients at the centre in

Norway, an annual individual dose to the technician is less than 2–3 mSv, mostly

derived from administration of FDG to the patient. This is lower than the ICRP dose

limit of 20 mSv for the workers in nuclear industry(Seierstad et al, 2007).



Other than the studies discussed above, studies about the radiation dose imparted by

whole body PET-CT scan are scarce, and none have addressed the cancer risk

induced. On the other hand, there is a rapid increase in the use of PET-CT scan.

Therefore, it is imperative that radiation dose and the associated cancer risk are

evaluated for PET-CT scans.


6.3 Materials and methods

Whole body PET-CT scans using a 64-slice CT system (Discovery PET-CT system,

GE Healthcare Systems, Milwaukee, WI) were studied. The dose from the PET-CT

scan is composed of doses from CT scan and PET scan. To measure the organ doses

of CT scan, direct measurement of radiation dose was performed. To estimate the

organ doses of PET scan, the dose coefficients recommended by ICRP publication 80

(ICRP, 1998) was applied as discussed in Chapter 2 and 3.




                                         111
CHAPTER 6                                                                 BS HUANG


6.3.1 Organ-specific CT dose measurement and simulation

The radiation dose of CT scan to patients was measured with an Alderson-Rando

phantom equipped with TLD-100 chips, as introduced in Chpater 2. A total number

of 266 chips were distributed in various organs of interest in the Rando phantom and

6 chips were used for controls (Table 6.1). To measure the radiation dose of skin, the

chips were placed inside small black plastic bags which were then attached to the

skin surface at 14 different positions (forehead, eye, sternal notch, para umbilicus,

supra pubic, upper thoracic, lumbar spine and sacrum). The TLD positions of bone

surface and bone marrow were considered to be the same, as it is hard to identify the

two organs in this phantom. The several readings of a specific organ were averaged

to calculate the absorbed dose of this organ. The distribution of chips was shown in

the Table 6.1.



Three protocols for whole-body CT scan currently used on the patients in our unit

were studied (Table 6.2). Protocols A and B were identical in kVp, rotation time,

slice thickness and pitch factor, but differed in tube current with “automA”

technology applied to protocol A. Protocol C was higher in kVp and tube current

which also applied automA technology. With the automA technology, tube current is

adjusted according to the patient anatomy to a user-selected noise level. This reduce

radiation dose compared to fixed-tube current protocols because the tube current is

automatically reduced for smaller patients or smaller anatomic regions.



In addition, CT dose was simulated using the well-established software ImPACT

(Medical Devices Agency, 2006), as described in Chapter 2. The results from

ImPACT simulation was used for verification in this part of our study.

                                         112
CHAPTER 6                                                                 BS HUANG


6.3.2 Organ-specific PET dose calculation

A whole-body emission PET scan of the same transverse coverage as the CT scan

was obtained with a 2-minute 45-second acquisition per bed position with the

scanner operating in the 3-dimensional mode. Typically, 5 or more bed positions

were obtained and the total scan time was for about 20 minutes. Organ-specific dose

were calculated according to the method introduced in Chapter 2. The average

activity of 18F-FDG administered in the unit was assumed to be 370 MBq for adults

(male and female).


6.3.3 Effective dose calculation

To calculate effective dose from PET-CT scan, the method introduced in Chapter 2

was applied. The effective dose was given by

                              Deff = ∑WT (Dorgan + Dorgan)
                                           CT       PET

                                    organ



where D eff is the overall effective dose, D organ and Dorgan are organ-specific doses
                                             CT         PET




of CT and PET respectively, WT is the tissue weighting factor for organ or tissue T as

listed in ICRP publication 103. The factors are for both genders.


6.3.4 LAR of cancer incidence estimation for US and the HK population

In Chapter 3 the tables for estimating the excessive lifetime cancer risk of US and

HK patients undergoing radiation exposure were set up (Table 3.7 and Table 3.8).

The LAR of cancer incidence of US and the HK population can be determined by

organ doses and data in Table 3.7 and Table 3.8 using the methods introduced in

Chapter 2. We repeated the analysis for a series of ages (from 20 to 80 years) and for

both genders.


                                            113
CHAPTER 6                                                             BS HUANG


6.3.5 LAR of cancer incidence compared to baseline

The baseline lifetime risk of cancer incidence for US and the HK population was

calculated according to the methods introduced in Chapter 2 (Table 2.9 and 2.10).

After calculating the LAR, the total cancer risk was acquired by summing the

baseline lifetime cancer risk and LAR, and then the proportion of LAR in total

cancer risk can be calculated.




                                      114
CHAPTER 6                                                                     BS HUANG


6.4 Results

6.4.1 Radiation doses

Measured organ doses and calculated effective doses from CT scan were summarized

in Table 6.3. According to the tissue weighting factors of ICRP publication 103, for

females, the effective doses of the three CT protocols were 7.22 mSv, 18.56 mSv,

25.68 mSv respectively; and for males were 7.42 mSv, 18.57 mSv, 25.95 mSv

respectively which were lower than that of females. The CT doses were different for

the three protocols. The doses of protocol A were higher than protocol B, and lower

than protocol C. This situation existed in every organ. The radiation doses to lens,

due to its deterministic effect, were also measured in our study, as shown in Table

6.3. Table 6.4, Table 6.5 and Table 6.6 shows the CT doses simulated using ImPACT

for the three protocols. For protocols A and C which applied automA technology, two

fixed tube currents (upper and lower limits of tube current) were input for dose

simulation as ImPACT requires fixed value. Hence a dose range was acquired. The

CT effective dose for protocol B calculated by ImPACT was 16.10 mSv and 16.40

mSv for females and males respectively. For the protocol A, the effective doses to

females for 100mA and 300mA were calculated to be 6.40 mSv and 19.10 mSv

respectively, and to males 6.60 mSv and 19.70 mSv respectively.


The calculated doses for PET scans were shown in Table 6.7. The effective dose

from the PET scan was 6.23 mSv. Doses from PET scan to gonads, bladder and

uterus were higher than the other organs, and were 5.00 mSv, 7.80 mSv, 59.20 mSv
                                                             18
respectively. This was due to the final accumulation of       F in the bladder. Other

organ doses range from 2.50 mSv to 4.80 mSv. PET doses of lens were not

calculated as there is no dose coefficient in ICRP publication 80 for lens.

                                          115
CHAPTER 6                                                                  BS HUANG




The total effective dose of the combined PET-CT scan was obtained by summing the

effective dose of CT and PET scan. Hence the total effective dose of the combined

PET-CT scan were 13.45 mSv, 24.79 mSv, 31.91 mSv for females and 13.65 mSv,

24.80 mSv, 32.18 mSv for males for protocols A, B and C respectively.


6.4.2 Cancer risks

6.4.2.1 Whole body cancer risk


LARs of cancer incidence induced from whole body PET-CT scan for the US and the

HK populations are illustrated in the Table 6.8 and Figure 6.1. The risks are

particularly high in younger ages and decrease with age. LARs for women are higher

than men across all ages. For example, for 20-year-old females in the US, the LARs

of cancer incidence are 0.231%, 0.403%, and 0.514% and for 20-year-old males in

the US, 0.163%, 0.254% and 0.323% respectively. For the HK population, the LARs

of cancer incidence are 0.267%, 0.484%, and 0.622% for 20-year-old females and

0.172%, 0.284% and 0.368% for 20-year-old males respectively. The risks for the

HK population are higher than for the US population for both genders and all ages,

as shown in Figure 6.1. For example, at age 20, the LARs of cancer incidence to the

HK population was 5.5% to 20.9% higher than the US population, and at age 80, the

incidence was 6.5% to 47.9% higher. The difference in risks between these two

populations is larger for females compared to males, at older ages and using higher

dose CT protocols. The largest difference is between risks to females using protocol

C (the difference range from 20.9% to 47.9% at age 80), and the smallest difference

is between risks to males using protocol A (the difference range from 2.3% to 6.5%).




                                         116
CHAPTER 6                                                                 BS HUANG


6.4.2.2 Comparison with baseline lifetime cancer incidence

According to the calculated baseline data in chapter 2, the proportion of LAR in total

cancer risk was calculated, as shown in Table 6.9. Due to the lower baseline risk but

higher LAR for females, the proportion that LAR contributed to total cancer risk to

females was higher than males, which suggested that the risk caused by radiation to

females was more severe than males. The proportion of LAR in total cancer

incidence is higher for the HK population than the US population, due to the lower

baseline cancer incidence.


6.5 Discussion

It is not surprising that dose differences were found among the three different CT

protocols which varied in tube current, tube potential and noise level (for automA

protocols). Generally, higher tube current and potential, and lower noise level

provide better quality CT images but impart a higher radiation dose and consequently,

a higher cancer risk. This underscores the fact that a balance between image quality

and radiation dose should be achieved in PET-CT scan protocols.



The CT dose measured was compared with the doses calculated using ImPACT for

verification. For protocol B, the effective dose calculated was 16.40 mSv for males

and 16.10 mSv for females, while the measured result were 18.56 mSv for males and

18.57 mSv for females. The simulated results were 12% (for males) and 13% (for

females) lower than measured results. The two results were concluded to be in good

accordance with each other, as it was reported that ImPACT underestimated CT

radiation dose by about 18% (Groves et al, 2004). For protocol A, the effective dose

of female for 100mA and 300mA was simulated to be 6.40 mSv and 19.10 mSv


                                         117
CHAPTER 6                                                                  BS HUANG


respectively, while the dose value measured for female was 7.22 mSv. For male the

calculated values using ImPACT were 6.60 and 19.70 mSv respectively, and the

measured dose was 7.42 mSv. The measured doses were within the range of the

calculated doses using ImPACT but were at the low end because a high noise level of

20 was selected for the scan. As it has been reported that ImPACT underestimates CT

radiation dose by about 18% (Groves et al, 2004), we conclude that the results from

ImPACT were in good accordance with the TLD measurements. Nevertheless, we

suggest that the ImPACT software should be updated for dose calculation.



The result of our research is comparable with the PET-CT radiation dose reported in

the literature. Some variability exists due to the differences in CT protocols, PET

dose estimating methods, and PET-CT scanners. A study in Japan (Ghotbi et al,

20070) estimated the total effective dose of a whole-body PET-CT scan to be

between 6.34 and 9.48 mSv for the average Japanese individual with a 60-kg body

weight. This lower dose was due to the use of a cancer screening CT protocol which

applied lower tube current and potential, although the transfer factor applied for PET

scan’s effective dose estimation was 2.1×10–2 mSv/MBq and was a little higher than

the factor we used. Our result was in good accordance with Brix et al, who measured

the effective dose per PET-CT examination for four different diagnostic CT protocols

and this ranged from 23.7 mSv to 26.4 mSv (Brix et al, 2005). Brix’s work applied

the same CT and PET dose estimating method as our presented study but the CT

scanners were 2-, 4- or 16-slice detectors. In another study, Wu et al reported

diagnostic CT (140kV, 80mA, 0.8 second, pitch 3) effective dose to be 18.97 mSv

using TLD and Rando phantom, and PET effective dose to be 10.72 mSv (Wu et al,

2005). These CT dose results are within the range of our measurement results but the


                                         118
CHAPTER 6                                                                    BS HUANG


coefficient for calculating the PET scan effective dose was 2.9×10–2 mSv/MBq,

according to the work of Deloar et al (Deloar et al, 1998).



We measured specifically the organ dose imparted to the lens by CT although this is

not relevant to effective dose calculation as the main risk from lens dose is the

deterministic effect (lens opacity and cataract formation), rather than the stochastic

effect (mainly cancer risk) (ICRP, 1991; ICRP, 2007). The calculated dose to the lens

was between 8 mSv and 27 mSv, and the higher end of the range exceeds the dose

limit of 15 mSv as recommended in ICRP publication 103 for the public. Hence

protection to the lens is recommended in whole body PET-CT scan.



Cancer risks calculated for the HK population were higher than the US population

across all ages, and especially for females, the older ages and higher dose CT

protocols. This is mainly due to the differences in life table and cancer statistics data

between the two populations. Firstly, Hong Kong-ers have a longer life expectancy

than Americans to develop cancer after radiation exposure. Secondly, the baseline

cancer incidences of the organs more sensitive to radiation are higher in Hong Kong

than US (Table 4.5). For example, the most prevalent cancer in Hong Kong is in the

lung which is given a high tissue weighting factor of 0.12 in ICRP publication 103,

while in the US it is in the prostate with a factor of 0.013.



There are inherent errors and limitations in our study, which is mainly from the dose

measurement (calculation) and cancer risk estimation methods applied in this study

as discussed in Chapter 2 and Chapter 3.




                                           119
CHAPTER 6                                                                  BS HUANG


The LAR of cancer incidence caused by whole body PET-CT scan was compared

with baseline cancer rate, as shown in Table 6.11 and Table 6.12. The proportion of

LAR in total cancer incidence is higher for the HK population than the US

population, which suggests that the risk induced by PET-CT scan is of greater impact

in the HK population.



As whole body PET-CT scan is accompanied by substantial radiation dose and

cancer risk, risk-benefit should be carefully weighed prior to every scan. This is

especially important when clinical utility is less well-established or anecdotal and

when used in younger patients. In the evaluation of patients with known cancer,

although cancer risks from radiation may be of less impact, the information is still of

interest and relevant to patient education. Moreover, cancer patients often have

multiple PET-CT examinations for response assessment and treatment monitoring,

and survival rates are markedly improved nowadays. From a public view-point, the

dose of up to 32 mSv per scan adding to the background radiation is non-negligible.

Therefore, protocols of PET-CT scan should be optimized for reducing dose and its

associated cancer risk.


6.6 Conclusion

Our results suggest that the excessive cancer risk induced by whole body PET-CT

scan is non-negligible, and higher for younger patients and women. Hence the

protocols of PET-CT scan should be optimized to reduce the dose and cancer risk.




                                         120
CHAPTER 6                                                                                                             BS HUANG


Figures:
                                            0.8%




                  LAR of Cancer Incidence
                                                                                                  Hong Kong Females
                                                                                                  American Females
                                            0.6%                                                  Hong Kong Males
                                                                                                  American Males
                                            0.4%


                                            0.2%


                                            0.0%
                                                      20    30          40         50        60   70        80
                                                                       Exposed Age (years)

                                                                 (a)        Protocol A
                                            0.8%                                                  Hong Kong Females
              LAR of Cancer Incidence




                                                                                                  American Females
                                            0.6%                                                  Hong Kong Males
                                                                                                  American Males

                                            0.4%


                                            0.2%


                                            0.0%
                                                   20      30      40      50       60            70        80
                                                                   Exposed Age (years)

                                                                 (b)        Protocol B
                                            0.8%                                                  Hong Kong Females
               LAR of Cancer Incidence




                                                                                                  American Females
                                            0.6%                                                  Hong Kong Males
                                                                                                  American Males

                                            0.4%


                                            0.2%


                                            0.0%
                                                   20      30          40       50       60       70        80
                                                                       Exposed Age (years)

                                                                 (c)        Protocol C

Figure 6.1:                                 LAR of cancer incidence induced by whole body PET-CT scan.
                                            Exposed age is the age when the person receives the PET-CT scan.


                                              Keys:
                                              LAR= lifetime attributable risk.



                                                                             121
CHAPTER 6                                                   BS HUANG



Tables:


                  organ              number of TLDs
                  background         6
                  brain              14
                  esophagus          10
                  thyroid            5
                  lung               61
                  thymus             2
                  heart              8
                  stomach            15
                  liver              24
                  spleen             7
                  adrenal            2
                  pancreas           4
                  kidney             12
                  colon              13
                  bone marrow        15
                  large intestines   9
                  small intestines   11
                  testicle           2
                  prostate           4
                  bladder            12
                  muscle             12
                  skin               28
                  total              272

Table 6.1:   Thermoluminescent dosimeters (TLD) distribution in Rando
             phantom.




                                     122
CHAPTER 6                                                             BS HUANG



                          U        t      Hcol     p     C           N
      CT scan protocol
                          (kVp)    (s)    (mm)           (mA)
             A            120      0.5    0.625    0.984 100~300     20
             B            120      0.5    0.625    0.984 250         -
             C            140      0.5    0.625    0.984 150~350     3.5

Table 6.2:   Parameters of CT protocol A, B and C.


                 Keys:
                 U=tube potential, t=rotation time, Hcol=slice thickness, p=pitch
                 factor, C=tube current or current range, N=noise level (for
                 automA technology).




                                         123
CHAPTER 6                                                              BS HUANG



                              protocol A      protocol B       protocol C
   organ              WT      female male     female male      female male
   gonad              0.08    8.8      8.43   18.95    20.43   27.04 27.43
   colon              0.12    7.77     8.36   22.10    18.63   30.97 29.28
   stomach            0.12    7.4      7.56   19.44    19.22   26.04 26.17
   lung               0.12    6.73     6.78   17.34    16.56   25.24 24.75
   bone
                      0.12    5.98    6.00    17.38   17.46    21.14   21.22
   marrow
   breast             0.12    5.73    7.33    13.89   18.98    19.12   25.27
   thyroid            0.04    10.61   10.06   27.25   27.13    37.47   36.39
   liver              0.04    8.45    8.02    20.87   19.77    29.88   28.33
   esophagus          0.04    7.63    7.45    19.79   18.72    28.32   27.21
   bladder            0.04    6.4     6.18    15.33   14.32    21.85   20.75
   skin               0.01    7.04    6.92    17.69   17.98    25.29   25.27
   bone
                      0.01    5.97    5.99    14.11   17.43    19.14   21.19
   surface
   brain              0.01    7.74    8.40    20.34   19.96    28.97   29.86
   kidney             0.013   8       7.90    20.16   19.45    28.08   27.41
   spleen             0.013   6.92    7.56    19.40   19.59    25.46   26.72
   adrenal            0.013   7.95    7.12    19.25   18.05    30.81   28.23
   uterus/prostate    0.013   7.33    7.44    19.28   17.35    27.11   25.86
   pancreas           0.013   7.23    7.07    18.99   17.73    27.05   25.84
   Small intestines   0.013   7.3     6.45    17.24   14.78    25.56   22.24
   large intestines   0.013   7.49    6.84    17.35   15.47    25.38   22.90
   thymus             0.013   7.49    7.17    17.28   18.48    23.99   24.24
   muscle             0.013   6.38    6.74    15.58   14.31    21.18   20.81
   heart              0.013   7.00    7.92    18.44   20.16    25.26   28.09
   lens                 -     8.1     8.3     18.4    18.6     27.2    27.3
   effective dose
                        1     7.22    7.42    18.56   18.57    25.68   25.95
   (mSv)

Table 6.3:    Measured dose for males and females (in mSv) from CT scans
              using thermoluminescent dosimeters (TLD). Effective doses were
              calculated using the tissue weighting factors (WT) recommended in
              ICRP publication 103.




                                       124
CHAPTER 6                                                        BS HUANG



                            tube current=100mA   tube current=300mA
     organ                  female     male      female    male
     gonads                 6.1        7.6       18        23
     bone marrow (red)      5.4        5.4       16        16
     colon                  6          6.1       18        18
     lung                   7.4        7.4       22        22
     stomach                6.9        6.9       21        21
     bladder                7.1        7.3       21        22
     breast                 5.5        5.5       17        17
     liver                  6.6        6.6       20        20
     esophagus (thymus)     8.1        8.1       24        24
     thyroid                11         11        32        32
     skin                   4.4        4.6       13        14
     bone surface           10         11        30        32
     effective dose (mSv)   6.40       6.60      19.10     19.70

Table 6.4:   Dose from CT scan protocol A, simulated using ImPACT
             spreadsheet. Effective doses were calculated using the tissue
             weighting factors (WT) recommended in ICRP publication 103.




                                   125
CHAPTER 6                                                        BS HUANG



                 organ                      female   male
                 gonads                     15       19
                 bone marrow (red)          14       14
                 colon                      15       15
                 lung                       19       19
                 stomach                    17       17
                 bladder                    18       18
                 breast                     14       14
                 liver                      16       16
                 esophagus (thymus)         20       20
                 thyroid                    27       27
                 skin                       12       12
                 bone surface               26       26
                 effective dose/mSv         16.10    16.40

Table 6.5:   Dose from CT scan protocol B, simulated using ImPACT
             spreadsheet. Effective doses were calculated using the tissue
             weighting factors (WT) recommended in ICRP publication 103.




                                      126
CHAPTER 6                                                        BS HUANG



                            tube current=150mA   tube current=350mA
     organ                  female      male     female      male
     gonads                 14          17       33          39
     bone marrow (red)      13          13       30          30
     colon                  14          14       33          33
     lung                   17          17       41          41
     stomach                16          16       38          38
     bladder                17          17       39          39
     breast                 14          14       32          32
     liver                  16          16       37          37
     esophagus (thymus)     18          18       42          42
     thyroid                23          23       54          54
     skin                   11          11       26          26
     bone surface           25          25       58          58
     effective dose (mSv)   15.10       15.40    35.40       35.90

Table 6.6:   Dose from CT scan protocol C, simulated using ImPACT
             spreadsheet. Effective doses were calculated using the tissue
             weighting factors (WT) recommended in ICRP publication 103.




                                   127
CHAPTER 6                                                              BS HUANG



       organ                FT (μGy/MBq)    organ dose/mSv     WT
       gonads               13.5            4.995              0.08
       lung                 10              3.7                0.12
       stomach              11              4.07               0.12
       colon                13              4.81               0.12
       bone marrow          11              4.07               0.12
       esophagus            11              4.07               0.04
       thyroid              10              3.7                0.04
       liver                11              4.07               0.04
       bladder              160             59.2               0.04
       breast               6.8             2.516              0.12
       bone surface         11              4.07               0.01
       skin                 8               2.96               0.01
       brain                11              4.07               0.01
       thymus               11              4.07               0.013
       spleen               11              4.07               0.013
       adrenal              11              4.07               0.013
       pancreas             11              4.07               0.013
       kidney               11              4.07               0.013
       large intestines     11              4.07               0.013
       small intestines 11                  4.07               0.013
       uterus (female)      21              7.77               0.013
       muscle               11              4.07               0.013
       heart                11              4.07               0.013
                effective dose (mSv)                    6.23

Table 6.7:   Calculated doses for males and females (in mSv) from PET scans
             using dose coefficients (FT) from ICRP publication 80. Effective
             doses were calculated using the tissue weighting factors (WT)
             recommended in ICRP publication 103.




                                      128
CHAPTER 6                                                         BS HUANG



                                 males’ LAR           females’ LAR
       exposed age (years)      HK         US        HK          US
               20            0.172%     0.163%    0.267%      0.231%
               30            0.119%     0.114%    0.177%      0.150%
               40            0.114%     0.110%    0.164%      0.132%
               50            0.104%     0.101%    0.145%      0.115%
               60            0.086%     0.084%    0.121%      0.094%
               70            0.060%     0.057%    0.088%      0.066%
               80            0.029%     0.027%    0.048%      0.033%
                 (a) LAR of PET-CT scan with CT protocol A.

                                 males’ LAR           females’ LAR
       exposed age (years)      HK         US        HK          US
               20            0.284%     0.254%    0.484%      0.403%
               30            0.195%     0.176%    0.316%      0.257%
               40            0.185%     0.166%    0.289%      0.220%
               50            0.167%     0.152%    0.252%      0.188%
               60            0.137%     0.124%    0.192%      0.152%
               70            0.093%     0.084%    0.148%      0.106%
               80            0.044%     0.039%    0.078%      0.053%
                (b) LAR of PET-CT scan with CT protocol B.

                                 males’ LAR           females’ LAR
       exposed age (years)      HK         US        HK          US
               20            0.368%     0.323%    0.622%      0.514%
               30            0.252%     0.222%    0.405%      0.327%
               40            0.239%     0.210%    0.370%      0.278%
               50            0.215%     0.192%    0.322%      0.237%
               60            0.176%     0.156%    0.264%      0.190%
               70            0.120%     0.105%    0.187%      0.132%
               80            0.057%     0.049%    0.098%      0.066%
                 (c) LAR of PET-CT scan with CT protocol C.

Table 6.8:   LAR of cancer incidence induced by whole body PET-CT scan.


               Keys:
               LAR= lifetime attributable risk.




                                     129
CHAPTER 6                                                             BS HUANG



                        protocol A       protocol B           protocol C
      age(years)
                     female male female          male     female      male
             20       1.14% 0.52% 2.04% 0.86%              2.61%     1.11%
             30       0.76% 0.36% 1.36% 0.59%              1.73%     0.77%
             40       0.73% 0.35% 1.29% 0.57%              1.64%     0.74%
             50       0.72% 0.33% 1.24% 0.54%              1.57%     0.69%
             60       0.73% 0.31% 1.15% 0.49%              1.58%     0.63%
             70       0.83% 0.32% 1.38% 0.49%              1.73%     0.63%
               (a) proportion of LAR in total cancer incidence (HK).

                       protocol A        protocol B           protocol C
     age(years)
                    female    male    female     male      female     male
         20         0.609% 0.365% 1.059% 0.569% 1.348% 0.721%
         30         0.399% 0.254% 0.681% 0.390% 0.863% 0.493%
         40         0.356% 0.242% 0.592% 0.367% 0.747% 0.463%
         50         0.324% 0.223% 0.529% 0.334% 0.665% 0.421%
         60         0.292% 0.191% 0.470% 0.283% 0.590% 0.355%
         70         0.252% 0.152% 0.402% 0.222% 0.503% 0.278%
               (b) proportion of LAR in total cancer incidence (US).

Table 6.9:      The proportion of LAR induced by whole body PET-CT scan in
                total cancer incidence. The total cancer incidence is acquired by
                adding the LAR to baseline lifetime cancer incidence (refer to
                section 2.6 in Chapter 2)


                 Keys:
                 LAR= lifetime attributable risk.




                                        130
CHAPTER 7                                                                    BS HUANG




CHAPTER 7 CONCLUSION SUMMARY AND FUTURE

STUDIES


7.1 Conclusion summary

In our study, the radiation doses from cardiac CT scan and PET-CT scan were

measured or calculated, and the cancer risks associated with these doses were

estimated. The organ doses from CT scan were measured using humanoid phantoms

and TLDs, and were also simulated using dosimetry calculator (ImPACT). Dose

coefficients were selected to estimate the internal dose from PET scan. Tables of

LAR of cancer incidence for US and the HK population were updated or set up, and

used for estimating the cancer risk induced by cardiac CT scan or PET-CT scan.



Our results showed that the doses from the cardiac CT scan or PET-CT scan were

high compared to other CT applications (such as head CT, abdomen CT). Compared

to the baseline cancer incidence, the risks were considered to be high as it

contributed to up to 3.5% in total lifetime cancer incidence.



It was also found that the radiation dose induced cancer risks were higher in the HK

population than in the US population, in terms of both absolute LAR and proportion

of LAR in total lifetime cancer incidence. This is due to the different cancer statistics

and life span data in HK and in US.



The risks were higher in the younger population than in the older population, and


                                          131
CHAPTER 7                                                                  BS HUANG


especially high in children. It is because children have a longer life expectancy and

are more sensitive to radiation. Therefore, scanning of children should be carefully

justified and the scanning parameters should be tailored for them, according to the

ALARA rules.



From the comparison with other published results, it was concluded that the doses

and cancer risks can be reduced greatly by adjusting scanning parameters and using

dose-reduction techniques (prospectively ECG-gated scan, or ECG-modulated tube

current technology).


7.2 Future studies

As discussed in Chapter 2, there are substantial errors with the dose coefficients for

estimating the dose of PET scan. Further research into the kinetic modeling of
18
     F-FDG in human body would be necessary for more accurate calculation of the

dose coefficients and estimation of the dose.



In our study the cancer risks induced by internal radiation (from PET scan) were

estimated using the same methods as for external radiation (from CT scan), which

may be associated with uncertainties. Longitudinal studies should be performed for

internal radiation induced risk analysis and modeling, as internal radiation especially

from the radionuclides with long half life and long biological retention times may be

even more harmful (ICRP, 1991; ICRP, 2007).



Our study is based on the studies about radiation induced risk transport from

Japanese to the US population, and is limited by the fact that the risk transport


                                         132
CHAPTER 7                                                                 BS HUANG


analysis has not been studied for the HK population. To increase the accuracy of

cancer risk estimation for the Hong Kong population, further research into risk

transport analysis from the Japan population to the HK population should be

performed.



These LAR tables for estimating cancer incidence associated with radiation exposure

can be used to estimate not only the risks from medical radiation, but also from other

radiation sources. Future studies may be focused on estimating the cancer mortality

risk for the HK population.




                                         133
REFERENCE LIST                                                                              BS HUANG



                                             Reference List


A.R.Lakshmanan. 1991. How to Measure the New ICRU Operational Quantities with Individual
Dosemeters. Radiation Protection Dosimetry 35 253-259.

Aroua,A., Burnand,B., Decka,I., Vader,J.P., Valley,J.F. 2002. Nation-wide survey on radiation doses in
diagnostic and interventional radiology in Switzerland in 1998. Health Phys 83 46-55.

Bar-Shalom,R. et al. 2003. Clinical performance of PET/CT in evaluation of cancer: additional value
for diagnostic imaging and patient management. J Nucl Med 44 1200-1209.

Beyer,T. et al. 2000. A combined PET/CT scanner for clinical oncology. J Nucl Med 41 1369-1379.

Blodgett,T.M., Meltzer,C.C., Townsend,D.W. 2007. PET/CT: form and function. Radiology 242
360-385.

Brenner,D., Elliston,C., Hall,E., Berdon,W. 2001. Estimated risks of radiation-induced fatal cancer
from pediatric CT. AJR Am J Roentgenol 176 289-296.

Brenner,D.J. et al. 2003. Cancer risks attributable to low doses of ionizing radiation: assessing what
we really know. Proc Natl Acad Sci U S A 100 13761-13766.

Brenner,D.J., Elliston,C.D. 2004. Estimated radiation risks potentially associated with full-body CT
screening. Radiology 232 735-738.

Brenner,D.J., Hall,E.J. 2007. Computed tomography--an increasing source of radiation exposure. N
Engl J Med 357 2277-2284.

Brix,G. et al. 2005. Radiation exposure of patients undergoing whole-body dual-modality 18F-FDG
PET/CT examinations. J Nucl Med 46 608-613.

Brugmans,M.J., Buijs,W.C., Geleijns,J., Lembrechts,J. 2002. Population exposure to diagnostic use of
ionizing radiation in The Netherlands. Health Phys 82 500-509.

Cardis,E. et al. 2005. Risk of cancer after low doses of ionising radiation: retrospective cohort study in
15 countries. BMJ 331 77-

Chodick,G., Ronckers,C.M., Shalev,V., Ron,E. 2007. Excess lifetime cancer mortality risk attributable
to radiation exposure from computed tomography examinations in children. Isr Med Assoc J 9
584-587.

Cohen,R.A., Frush,D.P., Donnelly,L.F. 2000. Data acquisition for pediatric CT angiography: problems
and solutions. Pediatr Radiol 30 813-822.

Deloar,H.M. et al. 1998. Estimation of absorbed dose for 2-[F-18]fluoro-2-deoxy-D-glucose using
whole-body positron emission tomography and magnetic resonance imaging. Eur J Nucl Med 25
565-574.

Desjardins,B., Kazerooni,E.A. 2004. ECG-gated cardiac CT. AJR Am J Roentgenol 182 993-1010.

Dewey,M., Hoffmann,H., Hamm,B. 2007. CT coronary angiography using 16 and 64 simultaneous
detector rows: intraindividual comparison. Rofo 179 581-586.

Diederich,S., Lenzen,H. 2000. Radiation exposure associated with imaging of the chest: comparison
of different radiographic and computed tomography techniques. Cancer 89 2457-2460.

Einstein,A.J., Henzlova,M.J., Rajagopalan,S. 2007. Estimating risk of cancer associated with radiation

                                                  134
REFERENCE LIST                                                                          BS HUANG

exposure from 64-slice computed tomography coronary angiography. JAMA 298 317-323.

Einstein,A.J. et al. 2008. Radiation dose and cancer risk estimates in 16-slice computed tomography
coronary angiography. J Nucl Cardiol 15 232-240.

EPA. 1994. Federal Radiation Protection Guidance for Exposure of the General Public. Proposed
Recommendations.

European Study Group of Radiologists and Physicists. 2008. European Guidelines on Quality Criteria
for Computed Tomography (EUR 16262 EN, May 1999). www drs dk/guidelines/ct/quality/index htm
(Feb):

Flohr,T.G. et al. 2006. First performance evaluation of a dual-source CT (DSCT) system. Eur Radiol
16 256-268.

Francone,M. et al. 2007. Noninvasive imaging of the coronary arteries using a 64-row multidetector
CT scanner: initial clinical experience and radiation dose concerns. Radiol Med (Torino) 112 31-46.

Frush,D.P., Donnelly,L.F., Rosen,N.S. 2003. Computed tomography and radiation risks: what
pediatric health care providers should know. Pediatrics 112 951-957.

Frush,D.P., Herlong,J.R. 2005. Pediatric thoracic CT angiography. Pediatr Radiol 35 11-25.

Ghotbi,N., Iwanaga,M., Ohtsuru,A., Ogawa,Y., Yamashita,S. 20070. Cancer screening with
whole-body PET/CT for healthy asymptomatic people in Japan: re-evaluation of its test validity and
radiation exposure. Asian Pac J Cancer Prev 8 93-97.

Ghotbi,N., Iwanaga,M., Ohtsuru,A., Ogawa,Y., Yamashita,S. 2007. Cancer screening with
whole-body PET/CT for healthy asymptomatic people in Japan: re-evaluation of its test validity and
radiation exposure. Asian Pac J Cancer Prev 8 93-97.

Gilkeson,R.C., Ciancibello,L., Zahka,K. 2003. Pictorial essay. Multidetector CT evaluation of
congenital heart disease in pediatric and adult patients. AJR Am J Roentgenol 180 973-980.

Goldman,L.W. 2007. Principles of CT and CT technology. J Nucl Med Technol 35 115-128.

Goldman,L.W. 2008b. Principles of CT: multislice CT. J Nucl Med Technol 36 57-68.

Goldman,L.W. 2008a. Principles of CT: multislice CT. J Nucl Med Technol 36 57-68.

Groves,A.M. et al. 2004. 16-detector multislice CT: dosimetry estimation by TLD measurement
compared with Monte Carlo simulation. Br J Radiol 77 662-665.

Halac,M. et al. 2007. Female breast surface radiation exposure during FDG PET/CT examinations.
Nucl Med Commun 28 924-928.

Hays,M.T., Segall,G.M. 1999. A mathematical model for the distribution of fluorodeoxyglucose in
humans. J Nucl Med 40 1358-1366.

Hays,M.T., Watson,E.E., Thomas,S.R., Stabin,M. 2002. MIRD dose estimate report no. 19: radiation
absorbed dose estimates from (18)F-FDG. J Nucl Med 43 210-214.

Hermann,F. et al. 2008. Reduction of radiation dose estimates in cardiac 64-slice CT angiography in
patients after coronary artery bypass graft surgery. Invest Radiol 43 253-260.

Herzog,C. et al. 2008. Pediatric cardiovascular CT angiography: radiation dose reduction using
automatic anatomic tube current modulation. AJR Am J Roentgenol 190 1232-1240.

Hirai,N. et al. 2008. Prospective versus retrospective ECG-gated 64-detector coronary CT


                                                135
REFERENCE LIST                                                                          BS HUANG

angiography: assessment of image quality, stenosis, and radiation dose. Radiology 248 424-430.

HK Census and Statistics Department. 2006. Hong Kong Life Table (1971-2006). HK Census and
Statistics Department Website (Nov):

Hohl,C. et al. 2006. Estimation of radiation exposure in low-dose multislice computed tomography of
the heart and comparison with a calculation program. Eur Radiol 16 1841-1846.

Hollingsworth,C.L. et al. 2007. Pediatric cardiac-gated CT angiography: assessment of radiation dose.
AJR Am J Roentgenol 189 12-18.

Hong Kong Cancer Registry. 2005. Hong Kong Cancer Statistics 2005. Hospital Authority of Hong
Kong Website (Nov):

Horiguchi,J. et al. 2008. Radiation dose, image quality, stenosis measurement, and CT densitometry
using ECG-triggered coronary 64-MDCT angiography: a phantom study. AJR Am J Roentgenol 190
315-320.

Hospital Authority of Hong Kong. 2008. Five Leading Causes of Death 1990 - 2005 in Hong Kong.
Hospital Authority home page (Jul):

Hounsfield,G.N. 1995. Computerized transverse axial scanning (tomography): Part I. Description of
system. 1973. Br J Radiol 68 H166-H172.

Huda,W. 2007. Radiation doses and risks in chest computed tomography examinations. Proc Am
Thorac Soc 4 316-320.

Huda,W. 2002. Effective doses to adult and pediatric patients. Pediatr Radiol 32 272-279.

Husmann,L. et al. 2008. Feasibility of low-dose coronary CT angiography: first experience with
prospective ECG-gating. Eur Heart J 29 191-197.

ICRP. 1991. 1990 Recommendations of the International Commission on Radiological Protection
[Publication 60].

ICRP. 1998. Radiation Dose to Patients from Radiopharmaceuticals [Publication 80].

ICRP. 2007. Draft recommendations of International Commission on Radiological Protection
[Publication 103].

Jones DG,S.PC. 1991. Survey of CT practice in the UK. Part 3: normalised organ doses calculated
using Monte Carlo techniques (NRPB-R250).

Jones,S.C. et al. 1982. The radiation dosimetry of 2 [F-18]fluoro-2-deoxy-D-glucose in man. J Nucl
Med 23 613-617.

Kalender,W.A., Wolf,H., Suess,C. 1999. Dose reduction in CT by anatomically adapted tube current
modulation. II. Phantom measurements. Med Phys 26 2248-2253.

Kinahan,P.E., Hasegawa,B.H., Beyer,T. 2003. X-ray-based attenuation correction for positron
emission tomography/computed tomography scanners. Semin Nucl Med 33 166-179.

Lembcke,A. et al. 2006. Pictorial review: electron beam computed tomography and multislice spiral
computed tomography for cardiac imaging. Eur J Radiol 57 356-367.

Leschka,S. et al. 2005. Accuracy of MSCT coronary angiography with 64-slice technology: first
experience. Eur Heart J 26 1482-1487.

Maria Ranogajec-Komor. 2003. Thermoluminescence Dosimetry - Application in Environmental


                                                136
REFERENCE LIST                                                                              BS HUANG

Monitoring. Radiation Safety Management 2 2-16.

Matthew Budoff, Jerold Shinbane. 2006. Cardiac CT Imaging:diagnosis of cardiovascular disease.

Mazonakis,M., Tzedakis,A., Damilakis,J., Gourtsoyiannis,N. 2007. Thyroid dose from common head
and neck CT examinations in children: is there an excess risk for thyroid cancer induction? Eur Radiol
17 1352-1357.

Medical Devices Agency. 2006. ImPACT CT Patient Dosimetry Calculator.

Mejia,A.A. et al. 1991. Estimation of absorbed doses in humans due to intravenous administration of
fluorine-18-fluorodeoxyglucose in PET studies. J Nucl Med 32 699-706.

Mollet,N.R. et al. 2005. High-resolution spiral computed tomography coronary angiography in
patients referred for diagnostic conventional coronary angiography. Circulation 112 2318-2323.

Nakanishi,T., Kayashima,Y., Inoue,R., Sumii,K., Gomyo,Y. 2005. Pitfalls in 16-detector row CT of
the coronary arteries. Radiographics 25 425-438.

National Institutes of Health. 2003. Report of the NCI-CDC Working Group to Revise the 1985 NIH
Radioepidemiological Tables. NIH Publication No. 03-5387.

National Research Council of the National Academies. 2006. Health Risks From Exposure to Low
Levels of Ionizing Radiation: BEIR VII Phase 2.

NCRP. 1993. Limitation of Exposure to Ionizing Radiation, NCRP Report No. 116.

Niven,E., Thompson,M., Nahmias,C. 2001. Absorbed dose to the adult male and female brain from
18F-fluorodeoxyglucose. Health Phys 80 62-66.

NRC BEIR Committee. 1990. Health effects of exposure to low levels of ionizing radiation: BEIR V.

NRC BEIR Committee. 2006. Health Risks From Exposure to Low Levels of Ionizing Radiation:
BEIR VII Phase 2.

NRC BEIR Committee. 1988. Health Effects of Radon and Other Internally Deposited Alpha-Emitters
(BEIR IV).

Pelosi,E. et al. 2004. Value of integrated PET/CT for lesion localisation in cancer patients: a
comparative study. Eur J Nucl Med Mol Imaging 31 932-939.

Pierce,D.A., Preston,D.L. 2000. Radiation-related cancer risks at low doses among atomic bomb
survivors. Radiat Res 154 178-186.

Prat-Gonzalez,S., Sanz,J., Garcia,M.J. 2008. Cardiac CT: indications and limitations. J Nucl Med
Technol 36 18-24.

Roberts,W.T., Bax,J.J., Davies,L.C. 2008. Cardiac CT and CT coronary angiography: technology and
application. Heart 94 781-792.

Schoenhagen,P. 2008. Back to the future: coronary CT angiography using prospective ECG triggering.
Eur Heart J 29 153-154.

Seierstad,T. et al. 2007. Doses to nuclear technicians in a dedicated PET/CT centre utilising 18F
fluorodeoxyglucose (FDG). Radiat Prot Dosimetry 123 246-249.

Shope,T.B., Gagne,R.M., Johnson,G.C. 1981. A method for describing the doses delivered by
transmission x-ray computed tomography. Med Phys 8 488-495.



                                                  137
REFERENCE LIST                                                                           BS HUANG

Shuman,W.P. et al. 2008. Prospective versus retrospective ECG gating for 64-detector CT of the
coronary arteries: comparison of image quality and patient radiation dose. Radiology 248 431-437.

Siegel,M.J. 2005. Pediatric CT angiography. Eur Radiol 15 Suppl 4 D32-D36.

Siemens Healthcare. 2008. Dual Source CT. Siemens Healthcare Global Website (Jul):

Slovis,T.L. 2002. CT and computed radiography: the pictures are great, but is the radiation dose
greater than required? AJR Am J Roentgenol 179 39-41.

Slovis,T.L. 2003. Children, computed tomography radiation dose, and the As Low As Reasonably
Achievable (ALARA) concept. Pediatrics 112 971-972.

Snyder,W.S., Fisher,H.L., Jr., Ford,M.R., Warner,G.G. 1969. Estimates of absorbed fractions for
monoenergetic photon sources uniformly distributed in various organs of a heterogeneous phantom. J
Nucl Med Suppl-52.

Stolzmann,P. et al. 2008. Radiation dose estimates in dual-source computed tomography coronary
angiography. Eur Radiol 18 592-599.

Thomas,K.E. et al. 2006. Assessment of radiation dose awareness among pediatricians. Pediatr Radiol
36 823-832.

Trabold,T. et al. 2003. Estimation of radiation exposure in 16-detector row computed tomography of
the heart with retrospective ECG-gating. Rofo 175 1051-1055.

Tubiana,M. et al. 2008. Computed Tomography and Radiation Exposure. N Engl J Med 358 850-853.

United Nations Scientific Committee on the Effects of Atomic Radiation. 2000. Sources and Effects of
Ionizing Radiation: United Nations Scientific Committee on the Effects of Atomic Radiation;
UNSCEAR 2000 Report to the eneral Assembly, With Scientific Annexes.

UNSCEAR. 1988. Sources, Effects and Risks of Ionizing Radiation: The 1988 Report to the General
Assembly with Annexes.

UNSCEAR. 2000a. Sources and Effects of Ionizing Radiation: United Nations Scientific Committee
on the Effects of Atomic Radiation; UNSCEAR 2000 Report to the eneral Assembly, With Scientific
Annexes.

UNSCEAR. 2000b. Sources and Effects of Ionizing Radiation: United Nations Scientific Committee
on the Effects of Atomic Radiation; UNSCEAR 2000 Report to the eneral Assembly, With Scientific
Annexes.

Upton AC, Adelstein SJ, Brenner DJ, et al. 2001. Report No.136 Evaluation of the
Linear-Nonthreshold Dose-Response Model for Ionizing Radiation.

US Disaster Center. 2006. US Life Table 2005. Web Site of US Disaster Centre (Jul):

US National Cancer Institute. 2006a. US Cancer Statistics 1975-2005. Web Site of Surveillance
Research Program, NCI (Jula):

US National Cancer Institute. 2006b. US Cancer Statistics 2001-2005. Web Site of Surveillance
Research Program, NCI (Julb):

Valentin J. 2005. Low-dose Extrapolation of Radiation Related Cancer Risk. Ann ICRP 35 1-140.

Wagner,G.S., Batey,S.E., Mosleh-Shirazi,M.A. 2000. Directionality of extruded lithium fluoride
thermoluminescent dosemeters in a cobalt-60 beam. Br J Radiol 73 1007-1009.



                                                 138
REFERENCE LIST                                                                           BS HUANG

Ware,D.E., Huda,W., Mergo,P.J., Litwiller,A.L. 1999. Radiation effective doses to patients undergoing
abdominal CT examinations. Radiology 210 645-650.

Westra,S.J. et al. 1999. Three-dimensional helical CT of pulmonary arteries in infants and children
with congenital heart disease. AJR Am J Roentgenol 173 109-115.

Wintersperger,B.J., Nikolaou,K. 2005. Basics of cardiac MDCT: techniques and contrast application.
Eur Radiol 15 Suppl 2 B2-B9.

Wu,T.H. et al. 2005. A positron emission tomography/computed tomography (PET/CT) acquisition
protocol for CT radiation dose optimization. Nucl Med Commun 26 323-330.




                                                 139
LIST OF FIGURES                                                                BS HUANG




LIST OF FIGURES


Figure 1.1:   Three-dimension coronary CT angiography image captured with the GE
              Healthcare 64-slice LightSpeed VCT scanner.

Figure 1.2:   CT, PET and PET-CT images.

Figure 1.3:   ECG-modulated tube current technique.

Figure 2.1:   Rando phantom

Figure 2.2:   Thermoluminescent dosimeters (TLD-100).

Figure 2.3:   Pediatric phantom.

Figure 2.4:   Process of CT dose measurement using TLD.

Figure 2.5:   Annealing oven for pre-heating TLD chips (PTW-Freiburg, Freiburg, Germany).

Figure 2.6:   TLD reader (Harshaw, model QS5500, OH, USA).

Figure 2.7:   CT dose calculation using ImPACT spreadsheet.

Figure 4.1:   LAR of cancer incidence induced by pediatric cardiac CT scan.

Figure 4.2:   The proportion of LAR induced by pediatric cardiac CT scan in total cancer
              incidence.

Figure 5.1:   LAR of cancer incidence induced by coronary CT angiography.

Figure 5.2:   The proportion of LAR induced by coronary CT angiography in total cancer
              incidence.

Figure 6.1:   Lifetime attributable risk of cancer incidence induced by whole body PET-CT
              scan.




                                            140
LIST OF TABLES                                                                       BS HUANG




LIST OF TABLES


Table 1.1:    Comparison of temporal and spatial resolution of various CT scanners.

Table 1.2:    Typical organ doses from various radiologic studies.

Table 2.1:    Thermoluminescent dosimeters (TLD-100) calibration.

Table 2.2:    PET dose coefficients recommended in ICRP publication 80

Table 2.3:    PET dose coefficients recommended in MIRD report

Table 2.4:    Tissue weighting factors recommended in ICRP publication 60 and publication
              103

Table 2.5:    Table of LAR of cancer incidence for US population.

Table 2.6:    Examples of organ-specific LAR calculation (for 20-year-old females)

Table 2.7:    Example of composite dose calculation (for females).

Table 2.8:    Example of baseline lifetime risk of cancer incidence calculation

Table 2.9:    Baseline lifetime risk of cancer incidence in the HK population

Table 2.10:   Baseline lifetime risk of cancer incidence in the US population.

Table 3.1:    Example of LAR calculation using ERR model (male).

Table 3.2:    Example of LAR calculation using ERR model (female)

Table 3.3:    Summary of LAR result using ERR model

Table 3.4:    Example of LAR calculation using EAR model (male)

Table 3.5:    Example of LAR calculation using EAR model (female)

Table 3.6:    Summary of LAR results calculated by using EAR model

Table 3.7:    Table of LAR of cancer incidence for HK population.

Table 3.8:    Updated Table of LAR of cancer incidence for US population.

Table 4.1:    Thermoluminescent dosimeters (TLD) distribution in pediatric phantom.

Table 4.2:    CT doses for boys and girls (in mSv) from pediatric cardiac CT scan.

Table 4.3:    LAR of cancer incidence induced by pediatric cardiac CT scan.

Table 4.4:    The proportion of LAR induced by pediatric cardiac CT scan in total cancer
              incidence.

Table 4.5:    Top five cancer sites of US and HK populations.

                                              141
LIST OF TABLES                                                                  BS HUANG



Table 5.1:   Typical cancer risks from various radiologic studies.

Table 5.2:   Doses from coronary CT angiography simulated using ImPACT.

Table 5.3:   LAR of cancer incidence induced by coronary CT angiography.

Table 6.1:   Thermoluminescent dosimeters (TLD) distribution in Rando phantom.

Table 6.2:   Parameters of CT protocol A, B and C.

Table 6.3:   Measured dose for males and females (in mSv) from CT scans using TLD.

Table 6.4:   Dose from CT scan protocol A simulated using ImPACT spreadsheet.

Table 6.5:   Dose from CT scan protocol B simulated using ImPACT spreadsheet.

Table 6.6:   Dose from CT scan protocol C simulated using ImPACT spreadsheet.

Table 6.7:   Calculated doses for males and females (in mSv) from PET scans

Table 6.8:   LAR of cancer incidence induced by whole body PET-CT scan.

Table 6.9:   Proportion of LAR induced by whole body PET-CT scan in total cancer incidence.




                                             142
BIBLIOGRAPHY OF THE AUTHOR                                          BS HUANG




PUBLICATIONS AND PRESENTATIONS RELATED TO

THE THESIS


1. Huang BS, Law MWM, Mak HKF, Kwok SPF, Khong PL (2009). Pediatric CT
   coronary angiography using 64-slice CT with ECG-modulated tube current:
   radiation dose and cancer risk. American Journal of Roentgenology. Accepted
   for publication.

2. BS Huang, M Law, PL Khong (2008). WHOLE BODY PET-CT SCAN:
   ESTIMATION OF RADIATION DOSE AND CANCER RISK. Radiology.
   Accepted for publication.

3. BS Huang, JY Li, PL Khong, Y Shen, M Law (2008). RADIATION DOSE
   AND CANCER RISK ASSOCIATED WITH 64-MDCT CORONARY
   ANGIOGRAPHY (Oral Presentation). RSNA 94th Scientific Assembly and
   Annual Meeting, Chicago, U.S.A

4. Huang, B.; Law M.; Kwok, S.; Khong, P. (2008). EVALUATION OF
   RADIATION DOSE AND CANCER RISK TO HONG KONG CHINESE
   INDUCED BY ADULT WHOLE BODY PET-CT SCANS (Oral Presentation).
   ARRS 2008 Annual Meeting, Washington DC, U.S.A

5. Bingsheng Huang, Martin Wai-Ming LAW, Stephen P.F Kwok, Pek-Lan
   Khong(2007). EVALUATION OF RADIATION DOSE IN ADULT WHOLE
   BODY PET-CT SCANNING. 2nd Joint Scientific Meeting of The Royal
   College of Radiologists & Hong Kong College of Radiologists and 15th Annual
   Scientific Meeting of Hong Kong College of Radiologists, Hong Kong




                                     143

				
DOCUMENT INFO
Shared By:
Categories:
Tags:
Stats:
views:51
posted:7/25/2011
language:English
pages:154