Anwar Padhani MD
Mount Vernon Hospital Cancer Centre
London, UK
Monitoring bony metastases response with diffusion MRI
Objectives
�� To illustrate the potential of whole body DWI in the therapy resTo illustrate the potential of whole body DWI in the therapy response ponse
assessment of bone metastasesassessment of bone metastases
�� To illustrate that there are 4 patterns of response when signal To illustrate that there are 4 patterns of response when signal
intensity and ADC values are evaluated intensity and ADC values are evaluated
�� To provide advice on how to use WBTo provide advice on how to use WB--DWI in conjunction with DWI in conjunction with
anatomic MRI to successfully gauge bone marrow therapy responseanatomic MRI to successfully gauge bone marrow therapy response
• Kwee TC, et al. Whole-body diffusion-weighted MRI. Eur J Radiol 2009; 70:409–417
• Padhani AD, Koh DM, Collins D. Whole body diffusion MR imaging in cancer – current status and
research directions. Radiology – in press
•Padhani AR & Koh DM. Diffusion MRI for monitoring of treatment response. MRI Clin N Am 2011; 19:
181-209
• Padhani AR & Gogbashian A. Bony metastases - assessing response to therapy with whole body
diffusion MRI. Cancer Imaging 2011; - in press
Monitoring bone therapy response
�� Symptoms assessments (analgesic requirements) and development ofSymptoms assessments (analgesic requirements) and development of
skeletal related events are markers of therapeutic efficacy in cskeletal related events are markers of therapeutic efficacy in clinical trials linical trials
�� Serum markers are not useful for therapy response for the majoriSerum markers are not useful for therapy response for the majority of ty of
tumors that metastasize to bonetumors that metastasize to bone
–– Serum PSA: not reliable in late stage hormoneSerum PSA: not reliable in late stage hormone--refractory prostate disease; refractory prostate disease;
flare phenomenon in responding patients*flare phenomenon in responding patients*
–– CA15CA15--3: moderate sensitivity for metastatic disease (603: moderate sensitivity for metastatic disease (60--65%); flare reaction in 65%); flare reaction in
responding breast cancer patients**responding breast cancer patients**
�� Serum/urinary markers of osteoblastic and osteoclastic activity Serum/urinary markers of osteoblastic and osteoclastic activity monitor monitor
bone response bone response NOTNOT tumor responsetumor response
�� Circulating tumor cells (CTCs) are emerging as strong response bCirculating tumor cells (CTCs) are emerging as strong response biomarkers iomarkers
for breast, colorectal and prostate cancersfor breast, colorectal and prostate cancers
**ScherScher HI et al. JCO 2008; 26:1148HI et al. JCO 2008; 26:1148--5959
**Duffy MJ. **Duffy MJ. ClinClin ChemChem 2006; 52:3452006; 52:345--5151
Evaluating bony disease by MRI: combining sequences into whole body examinations
� T1-weighted spin-echo, T2-weighted (FS) & STIR– Cellular and water content
� T2*-weighted GRE (IP/OP)– Fat:water ratio and susceptibility induced by trabecular bone
� Dynamic contrast enhanced MRI– Vascularisation
� Ultrashort TE (UTE)– Trabecular bone structure
� DW-MRI– Perfusion, cell density, fat and water content
* * Schaefer JF, et al. TotalSchaefer JF, et al. Total--body MRbody MR--imaging in oncology. imaging in oncology. EurEur RadiolRadiol. 2006; 16(9):2000. 2006; 16(9):2000--15.15.
* Schmidt GP, et al. Whole* Schmidt GP, et al. Whole--body MRI for the staging and followbody MRI for the staging and follow--up of patients with metastasis. up of patients with metastasis. EurEur
J J RadiolRadiol 2009; 70: 3932009; 70: 393––400400
Whole body DWI for metastasis detection
�� Widespread availabilityWidespread availability
�� Better than CT scans and bone Better than CT scans and bone
scans for lytic bony diseasescans for lytic bony disease
�� AlsoAlso provides information about provides information about
soft tissue diseasesoft tissue disease
�� AdvantagesAdvantages
–– No ionizing radiationNo ionizing radiation
–– No injections of isotopes or contrast No injections of isotopes or contrast
mediummedium
–– Quick to perform & readQuick to perform & read
–– QuantitativeQuantitative
�� At a glance assessmentsAt a glance assessments
→→ Improves whole body MRI Improves whole body MRI
performanceperformance
Kwee TC, et al. WholeKwee TC, et al. Whole--body diffusionbody diffusion--weighted MRI. weighted MRI. EurEur J J RadiolRadiol 2009; 70: 4092009; 70: 409––417417
WB-DWI sub-protocol(STIR for FS – total 25mins)
CoilsCoils surfacesurface
FOV (cm)FOV (cm) 380380
Rectangular (%)Rectangular (%) 100100
Acquisition matrixAcquisition matrix 128i128i
Scan %Scan % 8080
SlicesSlices 5050
Thickness/gapThickness/gap 5/05/0
ZZ--axis coverage (cm)axis coverage (cm) 2525
TR (ms)TR (ms) 90009000
TE (ms)TE (ms) min ( 67)min ( 67)
TI (ms)TI (ms) 180180
Half scan factorHalf scan factor 6/8i6/8i
bb--valuesvalues 50, 90050, 900
NENE 66
Bandwidth (Hz/Bandwidth (Hz/pixpix)) 16281628
Scan time/station (Scan time/station (minsmins)) 6.156.15
b50b50 b900b900
4 stations; 50 slices, 4 stations; 50 slices,
5mm; b50, b9005mm; b50, b900Scouts; T1W/STIRScouts; T1W/STIR
Whole bodyWhole body
Normal bone marrow pattern
� Normal adult BM distribution established by 25 yo
– Yellow bone marrow: 10-20% water, fat cells (↓SI)
– Red bone marrow: 40-60% water & more cells (↑SI)
� Variable red BM atrophy and trabecular bone loss
after 40 yo (>♀)
→ Increase in yellow bone marrow
Normal adult BM Normal adult BM
distributiondistribution
NonNon--linear (paradoxical) linear (paradoxical)
relationship between ADC & relationship between ADC &
bone marrow cellularitybone marrow cellularity
�� Yellow (fatty) marrow Yellow (fatty) marrow →→ low low
SI & ADCSI & ADC
-- Low water content & cellularityLow water content & cellularity
-- Fat acts as a barrier to water Fat acts as a barrier to water
diffusion repelling waterdiffusion repelling water
-- Low perfusionLow perfusion
�� Red bone marrow Red bone marrow →→ higher SI higher SI
& ADC & ADC
-- More cells and waterMore cells and water
-- Less big fat cells & more small cellsLess big fat cells & more small cells
-- Higher perfusionHigher perfusion
�� TumorTumor & BM hyperplasia& BM hyperplasia →→highest SI but variable ADC highest SI but variable ADC
-- Highest water content & cellularity Highest water content & cellularity
within restricted bone marrow spacewithin restricted bone marrow space
-- Highest perfusionHighest perfusion
Assessing WBAssessing WB--DWI response to RxDWI response to Rx
�� Changes in tumor burden/volumeChanges in tumor burden/volume
�� Changes in signal intensity on high bChanges in signal intensity on high b--value imagesvalue images
–– Reflects changes in water diffusivity Reflects changes in water diffusivity andand water contentwater content
�� Changes in ADC valuesChanges in ADC values
–– Objective evaluations of water diffusivityObjective evaluations of water diffusivity
–– Amenable to numerical analysis Amenable to numerical analysis
–– Heterogeneity assessments are made possibleHeterogeneity assessments are made possible
•• Histogram & pixel map analyses (parametric response maps)Histogram & pixel map analyses (parametric response maps)
Hamstra DA, et al, J Clin Oncol 2007: 25:4104-4109
Biological processes involved in therapy induced changes in DWI
25/Mar/10 2/Dec/10
Multiple Myelomacyclophosphamide, dexamethasone, thalidomide (CDT) and bortezomib
Prior
Current
WBWB--DWIDWI
Metastatic bone disease: Metastatic bone disease:
not responding to therapynot responding to therapy
�� New areas of focal abnormal signal intensityNew areas of focal abnormal signal intensity
�� Increasing extent (becoming confluent) and intensity of Increasing extent (becoming confluent) and intensity of
lesionslesions
�� For BM disease ADC values can For BM disease ADC values can ↓↓, , ↔↔ or or ↓↓(heterogeneous; but (heterogeneous; but
remains below thresholds)!remains below thresholds)!
Signal intensity changes in nonSignal intensity changes in non--responseresponse
47 47 yoyo F F –– metastatic breast cancer 1metastatic breast cancer 1stst Taxanes/APD then Taxanes/APD then LetrozoleLetrozole
1111--33--1010 88--66--1010 88--99--1010
b800b800 b800b800 b800b800
Increases in Increases in
tumour volume & tumour volume &
ADC values with ADC values with
progressionprogression
�� 51 51 yoyo F metastatic F metastatic
breast cancerbreast cancer
�� Bone and liver diseaseBone and liver disease
�� Pre and post Taxanes/ Pre and post Taxanes/
APD/ Herceptin APD/ Herceptin
�� Bone tumor volume Bone tumor volume
29.3 29.3 →→ 179.2 cm179.2 cm33
�� ADC 780 ADC 780 →→ 930 930 μμmm22/s/s
DWDW--MRI and cellularity in bone marrow MRI and cellularity in bone marrow A
DC
AD
C
Signal intensity on high bSignal intensity on high b--value imagesvalue images
Dense malignant infiltration>80% H20
Yellow marrowYellow marrow
10% H10% H2200Hyperplastic/moderate
tumour infiltration60-80% H20
Red marrowRed marrow
4040--60% H60% H2200
NonNon--
effective Rxeffective Rx
Disease progression can cause modest increases, stable or slightDisease progression can cause modest increases, stable or slight
decreases in ADCdecreases in ADC
BM hyperplasia: increased SI but stable ADC40 F – post-partum, left mastectomy and axillary nodal dissection
PRIORPRIOR
CURRENTCURRENT30-July 2010 10-December-2010
ADC 920 ADC 920 ±± 266 266 →→ 920 920 ±± 211 211 µµmm22/s/s
Rx adjuvant chemotherapy with G-CSF support
Myeloma progressionMyeloma progression
↑↑ SI & SI & ↓↓ADCADC
STIRSTIR T1WT1W
1111thth JanJan
11stst AprilApril
b800b800 ADCADC
1111thth JanJan
11stst AprilApril
1111--JanJan--1010 11--AprilApril--1010
62M Multiple Myeloma 62M Multiple Myeloma –– Rx BortezomibRx Bortezomib
1111--JanJan--1010 11--AprilApril--1010
900 900 µµmm22/s/s
725 725 µµmm22/s/s
Variable changes in ADC in nonVariable changes in ADC in non--
responders (stable + progression)responders (stable + progression)
�� 26 patients with 26 patients with
metastatic prostate metastatic prostate
cancercancer
�� ChemotherapyChemotherapy
�� 100 lesions 100 lesions
–– 33 in responders33 in responders
–– 59 in progressors 59 in progressors
–– 8 stable8 stable
�� Limits of reproducibility Limits of reproducibility
of normal BM of normal BM ±± 86 86 µµmm22/s/s
Messiou C, et al. Assessing response in bone metastases in prostMessiou C, et al. Assessing response in bone metastases in prostate cancer with diffusion ate cancer with diffusion
weighted MRI. weighted MRI. EurEur RadiolRadiol 2011 2011 –– ahead of printahead of print
Limits of
reproducibilityCh
an
ge
in
AD
C (
µm
2/s
)
Courtesy of C. Messiou, Royal Marsden Hospital, London
-600
-400
-200
0
200
400
600
800
1000
1200
Ch
an
ge in
AD
C (
um
2/s
)
Responders
Progressors
Stable patients
WBWB--DWIDWI
Bony diseaseBony disease
responding to therapyresponding to therapy
�� Decrease in volume of focal abnormal signal intensityDecrease in volume of focal abnormal signal intensity
�� Signal intensity changes generally decrease but occasionally Signal intensity changes generally decrease but occasionally
unchanged/increase if there is a strong inflammatory response unchanged/increase if there is a strong inflammatory response
�� ADC values : marked increasesADC values : marked increases
Breast cancer (triple negative)Breast cancer (triple negative)Rx Rx GemCarboGemCarbo + APD (bisphosphonate)+ APD (bisphosphonate)
0101--OctOct--20102010
T1WT1W T2W+FST2W+FS WBWB--DWIDWI
b900b900
2929--JulyJuly--20102010
T1WT1W T2W+FST2W+FS WBWB--DWIDWI
b900b900
Decreases in DW-MRI is due to a combination of changes in tissue water content and cellularity
�� 37F 37F –– perimenopausalperimenopausal
�� Triple negative, Triple negative, metastatic breast cancermetastatic breast cancer
�� Rx Rx –– GemCarboGemCarbo plus APDplus APD
�� Diffuse increase in ADC Diffuse increase in ADC valuesvalues
�� ADC 883 ADC 883 →→ 1905 1905 μμmm22/s/s
�� CutCut--off values: 650 & off values: 650 & 1500 1500 μμmm22/s/s
2929--JulyJuly--20102010 0101--OctOct--20102010
b900b900
ADCADC
Prior
Current
Breast cancer (triple negative)Breast cancer (triple negative)Rx Rx GemCarboGemCarbo + APD (bisphosphonate)+ APD (bisphosphonate)
AD
CA
DC
Signal intensity on high bSignal intensity on high b--value imagesvalue images
Dense malignant infiltration>80% H20
Yellow marrowYellow marrow
10% H10% H2200Hyperplastic/moderate
tumour infiltration60-80% H20
Red marrowRed marrow
4040--60% H60% H2200
NonNon--
effective Rxeffective Rx
Effective Rx
Decreasing cellularity
Breast cancerBreast cancer
Rx Rx Capecitabine (3 cycles) + APD(3 cycles) + APD
2121--JanJan--20112011
T1WT1W T2W+FST2W+FS b900b900
0404--AprilApril--20112011
T1WT1W T2W+FST2W+FS b900b900
Breast cancerBreast cancerRx Rx capcitabinecapcitabine + +
APDAPD
�� 41F 41F –– postmenopausalpostmenopausal
�� Metastatic breast cancer Metastatic breast cancer --asymptomatic (before and asymptomatic (before and during Rx)during Rx)
�� Rx Rx –– cepcitabinecepcitabine (3 cycles) plus (3 cycles) plus APDAPD
�� Signal intensity is unchanged but Signal intensity is unchanged but marked increase in ADC valuesmarked increase in ADC values
�� ADC ADC →→ mean 900 to 1500 mean 900 to 1500 μμmm22/s/s
�� CutCut--off values: 650 & 1500 off values: 650 & 1500 μμmm22/s/s
2121--JanJan--20112011 0404--AprilApril--20112011
b900b900
ADCADC
CurrentCurrentPriorPrior
AD
CA
DC
Signal intensity on high bSignal intensity on high b--value imagesvalue images
Yellow marrowYellow marrow
10% H10% H2200Dense malignant
infiltration>80% H20
Hyperplastic/moderate tumour infiltration
60-80% H20
Red marrowRed marrow
4040--60% H60% H2200
NonNon--
effective Rxeffective Rx
Effective Rx
Decreasing cellularity NecrosisNecrosis
(T2(T2--shine through)shine through)
Response with GemCarbo+bisphosphonates → relapse after stopping Rx
Progression with bisphosphonates → response with docetaxel + bisphosphonates
23/Jul/10 1/Oct/10 20/Apr/118/Feb/1115/Dec/10
5/Jan/10 6/Apr/10 30/Jul/10 28/Oct/10 1/Mar/11
Proposed response criteria of BM lesionsEARLY after starting cytotoxic Rx
SI on high b-
value images
ADC in relation
cut-off
Biological inference and
response assessment
↑ Most pixels <cut-off Persistent hypercellularity → no
evidence of response
↑ Most pixels >cut-off Necrosis, T2-shine through → evidence
of response
↓ Most pixels >cut-off Hypocellularity → evidence of response
↓ Most pixels <cut-off Possible sclerotic or fibrotic reaction →
indeterminate for response
ADC change needs to be judged in relation to cutADC change needs to be judged in relation to cut--off values defined from untreated patients off values defined from untreated patients
examined using the same imaging protocol. This is likely to be texamined using the same imaging protocol. This is likely to be tumor type dependent. umor type dependent.
Criteria may not apply to nonCriteria may not apply to non--cytotoxic therapies. The timeline for the applicability of thesecytotoxic therapies. The timeline for the applicability of these
criteria are undefined criteria are undefined
Padhani AR & Gogbashian A. Bony metastases Padhani AR & Gogbashian A. Bony metastases -- assessing response to therapy with assessing response to therapy with
whole body diffusion MRI. Cancer Imaging 2011; whole body diffusion MRI. Cancer Imaging 2011; -- in pressin press
Some outstanding R&D questionsSome outstanding R&D questions
�� It is not clear what proportion of tumor cells have to be killedIt is not clear what proportion of tumor cells have to be killed for for
ADC changes to become detectable by DWADC changes to become detectable by DW--MRIMRI
–– Does the mechanism of cell death affect DWDoes the mechanism of cell death affect DW--MRI appearances? MRI appearances?
�� What is the reproducibility of WBWhat is the reproducibility of WB--DWI ADC estimates (that is, how DWI ADC estimates (that is, how
much of measured change can be considered much of measured change can be considered ““realreal””??
–– What factors affect the measurement reproducibility?What factors affect the measurement reproducibility?
→→ Essential information for the development of WBEssential information for the development of WB--DWI as a DWI as a
pharmacodynamic biomarker for use in drug trialspharmacodynamic biomarker for use in drug trials
�� How much change in ADC results in patient benefit (improvements How much change in ADC results in patient benefit (improvements
in symptoms and survival) for the variety of available therapiesin symptoms and survival) for the variety of available therapies??
→→ Essential information for the development of WBEssential information for the development of WB--DWI as a tool for DWI as a tool for
personalized medicinepersonalized medicine
Padhani AR, Koh DM & Collins D. Padhani AR, Koh DM & Collins D. Whole body diffusion MRI in cancer Whole body diffusion MRI in cancer -- current status & current status &
research directionsresearch directions. Radiology . Radiology -- in pressin press
Take home pointsTake home points
�� WBWB--DWI is a SI based tool for DWI is a SI based tool for bone marrow lesion bone marrow lesion
detection and therapy responsedetection and therapy response
–– Lytic bone deposits are better seen than sclerotic lesionsLytic bone deposits are better seen than sclerotic lesions
�� NonNon--linear (paradoxical) relationship between ADC & bone linear (paradoxical) relationship between ADC & bone
marrow signal intensitymarrow signal intensity
–– Disease progression causes heterogeneous changes in ADCDisease progression causes heterogeneous changes in ADC
–– Response assessments Response assessments →→ larger ADC increases larger ADC increases
�� WBWB--DWI should always be interpreted with conventional DWI should always be interpreted with conventional
imaging findingsimaging findings
�� WBWB--DWI has not yet been proven to impact meaningful health DWI has not yet been proven to impact meaningful health
outcomes in patients with metastatic bone diseaseoutcomes in patients with metastatic bone disease
Padhani AR & Gogbashian A. Bony metastases Padhani AR & Gogbashian A. Bony metastases -- assessing response to therapy with assessing response to therapy with
whole body diffusion MRI. Cancer Imaging 2011; whole body diffusion MRI. Cancer Imaging 2011; -- in pressin press