90 Y-PET versus 90 Y-Bremsstrahlung SPECT S Walrand Cliniques universitaires saint-Luc, Brussels 90 Y bremsstrahlung scan Simon and Feitelberg 1967 90 Y TOF PET/CT Lhommel et al. 2009 First 90 Y microspheres injected intra-arterially imaging
90Y-PET versus 90Y-Bremsstrahlung SPECT
S Walrand
Cliniques universitaires saint-Luc, Brussels
198Au rays scan 90Y bremsstrahlung scan
Simon and Feitelberg 1967
90Y TOF PET/CT
Lhommel et al. 2009
First 90Y microspheres injected intra-arterially imaging
Ford KW Phys Rev 1955
Hypothetical level diagram of 89Zr
as deduced from the neighboring nucleus 89Y.
A first excited 0+ state is expected
because of the large mutual interaction of the (g9/2)2 configuration.
2 or e+ e-
2x10-4 e+ observed Johnson et al. Phys Rev 1955
Nickles et al. IEEE TNM 2004
90Y micro-Derenzo phantom imaged
with a micro-PET P4 scanner
Selwyn et al. ARI 2006
e+: 3.2x10-5
Milestones in 90Y positron emission studies
The decay of 90Y has a minor branch to the 0+ excited state[1], followed by an internal e+ e- creation which happens in 32 out of one million decays [2]. Consequently, 90Y PET scan was proposed in order to assess the biodistribution [3] of 90Y-labelled therapeutic agents. A 61-year-old woman was referred for …………………………….
1. Ford. Phys Rev 1955 2. Selwyn et al. Appl Radiat Isot 2007 3. Nickles et al. IEEE Nucl Sci Symp Record 2004
30min
The additional liver low dose CT performed for the 90Y PET/CT evaluation
could be seen as an unjustified additional irradiation of the patient.
90Y: 40000 ± 5000 mSv CT: 2 mSv
amazing epoch
Lhommel, ...., Walrand EJNMMI 2010
Feasibility of 90Y TOF-PET based dosimetry in liver metastasis therapy using SIR-spheres.
true
TOF-PET
Dosimetry by kernel dose convolution
Relative activity
0 1
liver
2 3 4
True
mGy/MBQ
6.5 20.7 34.9 52.1 65.8
TOF-PET
mGy/MBQ
11. 24.6 37.3 52.6 63.5
Err
(%)
72 18.6 6.8 1.0 -3.4
FDG TOF-PET (day 1) FDG TOF-PET (day 57) 90Y TOF-PET (day 15)
Mean SUV = 5.96 Max. SUV =12.0
Mean SUV = 2.86 Max. SUV = 6.41
Mean dose: 85.3 Gy Max. dose:153.0 Gy
Mean dose: 29.9 Gy Max. dose: 37.7 Gy
necrotic tumour
90Y: 45min acquisition for 2 bed positions, i.e. 27cm
responding
progressive
90Y TOF-PET based dosimetry in liver metastasis therapy using resin spheres.
Lhommel, ...., Walrand EJNMMI 2010
HEF x D (Gy)
0 50 100 150 200 250
su
rviv
al
frac
tio
n
0·01
0·02
0·05
0·1
0·2
0·5
1
2
5
10
20
R=0·96
me
tab
olic
ra
tio
at
wee
k 7
po
st
thera
py
0·01
0·1
1
10
R2 = 0.96
Tumour early response-dose relationship in 90Y liver SIRT
Hb level (g/dL)8 10 12 14 16
HE
F
0
1
2
R=0·94
Walrand et al. EJNMMIres 2012
van Elmbt, Vandenberghe, Walrand et al. PMB 2011
Comparison of different crystals in 90Y PET
GSO
BGO
LYSO
Gemini TF LYSO TOF
Gemini TF LYSO
non-TOF
Allegro GSO
Exact HR+ BGO
36min 24min 12min
high count rate phantom hot spheres (1.7GBq 90Y)
van Elmbt, Vandenberghe, Walrand et al. PMB 2011
Comparison of different crystals in 90Y PET
TOF + PSF
TOF
PSF
5 10 15 20 25 30 35 40
0
20
40
60
80
100
hot sphere diameter (mm)
reco
ve
ry (
%)
Willowson et al. [13]
TOF + RR van Elmbt et al. [11]
Sfactor Stabin et al. [14-1]
QUEST multicentric study summary (Willowson et al. EJNMMI 2015)
Philips Gemini TF
Siemens Biograph mCT
GE Discovery 710 90Y Sfactor
recovery < 90Y Sfactor → local energy deposition method
Comparison of different crystals in 90Y-DOTATOC PET
low count rate kidney MIRD P19 phantom
modeling a 1h PET acquisition
after a typical first 90Y-DOTATOC therapy cycle
Walrand et al. JNM 2011
spine: 2.5MBq-12ml plastic vial
in 5cm diameter cylinder
filled with K2HPO4 solution
modeling bone attenuation
Kidney 35MBq Spline 100MBq
7ml tumours: 11,25 MBq
CT Gemini TF LYSO
Exact HR+ BGO
missing activity (radioactivity of LYSO)
50% plexiglas 50% H2O
H2O
Relative deviation: -18% -3%
dosimetry after each cycle
to optimize the next one
PET <> brems-SPECT in 90Y-DOTATOC therapy
Fabbri et al. Canc Bioth 2015
90Y-based PET and SPECT imaging
in locoregional brain treatment for high-grade gliomas
Fabbri et al. EJNMMI 2012
PET can provide superior image quality compared with brems-SPECT
and may enable reliable detection of extra-articular 90Y activity
Barber et al. JMI&RO 2013
90Y PET <> brems-SPECT post synovectomy
bin
0 50 100 150
co
un
ts
0.1
1
10
100
1000
10000
measured profile
source profile
scc profile
e+ e- pair production in LSO (X>1MeV) ?
(van Elmbt-Walrand)
Open issues in liver 90Y PET
spurious extra-hepatic counts
90Y-PET
90Y-TOF PET
acquired on a PET Biograph (courtesy of Dr. Ahmadzadehfar)
90Y vial in air true coincidences radial profile
Gate-GEANT4 simulations did not show
any significant e+e- creation in LSO
(Strydhorst et al. Med Phys 2016)
increasing tail
Internal bremsstrahlung:
a forgotten but significant effect in 90Y SPECT and PET imaging.
(Walrand et al. EANM Vienna 2017)
Aston 1927
1.E-7
1.E-6
1.E-5
1.E-4
1.E-3
1.E-2
1.E-1
1.E+0
1.E+1
0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0 2.2 2.4
h2o brems G4
fit h2o brems
int brems [33]
int brems [28]
fit int brems
photo
n p
er
MeV
per
decay
E (M)
Venkataramaiah et al.1980
Cengiz et al.2004
Int. Bremss. not modeled in Geant4
effect discovery knowledge in nuclear medicine
e+ emission 1955 2004
int. bremss. 1927 2017
nucleus E field
β-
1.E-5
1.E-4
1.E-3
1.E-2
1.E-1
1.E+0
0 500 1000 1500
18F spectra
G4 511-Lu176
gauss 511keV
spectra fit
1.E-5
1.E-4
1.E-3
1.E-2
1.E-1
1.E+0
0 500 1000 1500
2.80E-03
gauss 511keV
fit
1.E-5
1.E-4
1.E-3
1.E-2
1.E-1
1.E+0
0 200 400 600 800
111In spectra
gauss 171keV
gauss 245keV
spectra f it
1.E-5
1.E-4
1.E-3
1.E-2
1.E-1
1.E+0
0 200 400 600 800
LYSO-PET 18F spectra
keV keV
arb
itra
ry u
nits
acquisition
171keV gauss
245keV gauss
fit
acquisition
G4 pileup
511keV gauss
fit
A B GSO-PET 111In spectra
acquisition
511keV gauss
fit
LSO-PET 18F spectra C
keV
Crystal scintillation exhibits long energy resolution tail
Internal bremsstrahlung:
a forgotten but significant effect in 90Y SPECT and PET imaging.
(Walrand et al. EANM Vienna 2017)
1.E-5
1.E+0
0 250 500 750
0.000
0.020
0.040
0.060
0.080
0.100
-45 -30 -15 0 15 30 45
acq
G4 int+h2o+gauss
G4 h2o+tail
G4 int+h2o+gauss
theory
cp
s p
er
bin
bin
GSO-PET 32P true cc rate radial profile A B
e-
>>250keV
bin
≈250 keV
32P vial
A
B
[450,650] keV
BUT
Internal bremsstrahlung:
a forgotten but significant effect in 90Y SPECT and PET imaging.
(Walrand et al. EANM Vienna 2017)
On the origin and on the reduction of spurious extrahepatic counts
observed in 90Y non-TOF PET imaging post radioembolization
(Walrand et al. EANM Vienna 2017)
Brems Lu176 LOR-TOF Brems Lu176 3D-RAMLA Brems Lu176 3D-RAMLA
Brems Lu176 3D-RAMLA Brems Lu176 LORDS-3D-RAMLA Brems Lu176 LOR-TOF
A B C
D E F
Spurious extrahepatic counts originate from randoms
not perfectly corrected in non-TOF PET
Gate-GEANT4 simulations with int+ext bremss.
198Au rays scan 90Y bremsstrahlung scan
Simon and Feitelberg 1967
0.00
0.01
0.02
0.03
0.04
0.05
0.06
0.07
0.08
50 100 150 200 250 300
Energy (keV)
Co
un
ts / k
eV
/ M
Bq
.s
Total counts
Primary photons
Scattered photons
Camera backscatter
Collimator scatter & penetration
Lead x-rays
Pb
[50,150]keV
geometrical
[50,150]keV
bac
k-sc
atte
rin
g
>150keV reduced using a medium energy or high energy collimator
90Y SPECT: contamination components
PET
3 10-5 decay emits positron
3D sensitivity: 2.5 10-2
0.8 10-6
1.5% decay emits bremsstrahlung X ray in [50,150] keV
Collimator sensitivity 10-4
1.5 10-6
SPECT
But + 9 10-6 scattered X rays
90Y SPECT: state of the art MC reconstruction
without scatter correction
with MC scatter correction Dewaraja et al. Med Phys 2017 Rault et al. Med Phys 2010
Germanium detector
Total 90Y bremss. in h2o
Not commercial available and collimator-camera dependent
The liver concentration of 0.34 MBq/mL used in the phantom was limited by radiation safety requirements of the regulatory authority at our institution.
Some radioprotection officers are imposing ALARA
can be a threat to imaging development (dead time, background)
and at the end, also a threat patient care
Gate 6.2 : Geant4 9.5-p01 + Geant4 9.6-p02 for atom de-excitations and fluorescence emissions
crystal
Lead housing
Pyrex lightguide
Mu-metal shielding
PMTs dynodes
90Y source in perpex
BGO camera based on Exact HR+ BGO-PMTs blocks + flat pinhole collimator
GE400 AC Crystal: NaI or BGO
90Y bremsstrahlung Monte Carlo modeling
Walrand et al. Front Oncol 2014
Pb
pri
mar
y
cam
era
bac
k-sc
atte
rin
g
NaI
PMT
elect. boards
β-
MC simulation of camera contaminating x-rays in bremsstrahlung imaging
signal >
scatter
scatter >
signal
E (keV)
0 100 200 300 400 500
Sig
na
l to
sca
tte
r ra
tio
0.1
1
10
100
MC simulation of the signal to scatter ratio
Pyrex
Walrand et al. Front Oncol 2014
Pb
pri
mar
y
BGO
PMT
elect. boards
β-
signal >
scatter
scatter >
signal
E (keV)
0 100 200 300 400 500
Sig
na
l to
sca
tte
r ra
tio
0.1
1
10
100
E (keV)
0 100 200 300 400 500
Sig
na
l to
sca
tte
r ra
tio
0.1
1
10
100
MC simulation of camera contaminating x-rays in bremsstrahlung imaging
MC simulation of the signal to scatter ratio
Walrand et al. Front Oncol 2014
signal >
scatter
scatter >
signal
E (keV)
0 100 200 300 400 500
Sig
na
l to
sca
tte
r ra
tio
0.1
1
10
100
Continuous energy tomography (CET)
Pb
pri
mar
y
BGO
PMT
elect. boards
β-
MC simulation of camera contaminating x-rays in bremsstrahlung imaging
MC simulation of the signal to scatter ratio
Walrand et al. Front Oncol 2014
90Y bremss. NaI-pinhole SPECT SIRT phantom evaluation
pinhole SPECT of longitudinally extended organ requires helical acquisition
helical non-circular acquisition, bed & camera manually moved
positions measured by 3 tape rulers 2 hours for the motions
medium energy thyroid pinhole: focal length too long
and 1/2 of the crystal area used
Walrand EJNMMI research 2011
Bremsstrahlung helical pinhole SPECT is superior to 90Y-TOF-PET
true 90Y TOF-PET helical PH SPECT
not commercially available
physicist patient
d
e
f a
36 min 3.6 min 1 min b c
90Y liver-SIRT
dosimetry in 3min helical PH SPECT
in catheterization room → optimization
Walrand EJNMMI research 2011
Perspective: dose optimization in catheterization room
Perspective: dose optimization in catheterization room
1/3 of maximal MAA based activity 3 min SPECT liver dosimetry if wrong delivery: catheter repositioning
or allowed activity to reach maximal safe liver dose
0.0
0.5
1.0
1.5
2.0
2.5
0.0 0.5 1.0 1.5 2.0 2.5
500ps-TOF & rand. est.
500ps-TOF & rand. sub.
activity/4 and 4x20min
GSO
L[Y]SO non-TOF
BGO
bremss. MEGP 2H SPECT ( T↑ S↑↑ SR=12mm)
100ps-TOF bremss. helical MEPH 2H SPECT ( S↓ versus MEGP)
936mm
van Elmbt ... PMB 2011
CT
kidney PRRT MIRD p19 Walrand ... JNM 2010
90Y activity (GBq in a liver lobe)
SNR
(2
0 m
in a
cqu
isit
ion
tim
e)
1936 mm
Walrand ... EJNMMI res 2011
true
90Y+L[Y]SO
Conclusions in 90Y imaging
PET supersedes commercial bremss. SPECT in all therapies
PET is usable for dosimetry purpose (using local energy deposition method)
TOF PET is preferable in post radioembolization delivery check
State of the art MC reconstruction bremss. SPECT approaches PET performance
Standard pinhole bremss. SPECT supersedes TOF-PET
and could be used in catheterization room for optimization
BGO (or GSO) pinhole bremss. SPECT is the optimal modality