Interim PET with emphasis on the effect of drugs What can we learn from animal studies?
L. Brepoels, S. Stroobants
Chemosensitivity
Chemoresistence
Responders vs nonresponders
Sufficient response?
Death
False positive
Symptoms
quantification
PET-detection
(radiotherapy)
chemo
chemo
chemo
chemo
chemo
chemo
False negative
Diagnosis +chemo
Number of malignant cells
Principles of response assessment
Time
Early response assessment in DLBCL after 7 days of treatment • Materials and methods – – – –
29 patients Newly diagnosed DLBCL Treatment with R-CHOP PET/CT after 7 days
Early response assessment in DLBCL after 7 days of treatment 29 patients ∟ 17 patients positive on early PET ∟ 2 refractory disease 2 relapsed (12 and 23mths) ∟ 13 disease free (follow-up 20 mths)
∟ 12 patients negative on early PET ∟ No relapses (21 mts) Visual: NPV=100%, PPV=24% Quantitative: NPV= 100%, PPV=29%
SUVmean_all = 8.55 SUVmax_all = 18.34 Volmet_all = 658 ml
SUVmean_all = 5.66 SUVmax_all = 12.47 Volmet_all = 361 ml
CR
CR
Despite a significant residual uptake on early PET, this patient obtained a complete remission at interim PET, and is still disease free after a follow-up of 29 months
30 20 10 0 -10 -20 -30 -40 -50 -60 -70 -80 -90 -100
Analysis based on tumor bulk after partial volume correction 50 40 30 20 10 0 -10 -20 -30 -40 -50 -60 -70 -80 -90 -100
Analysis based on the worst responding lesion after partial volume correction
∆ SUVmax_worst 30 20 10 0 -10 -20 -30 -40 -50 -60 -70 -80 -90 -100 -80
∆ SUVmean_all
∆ SUVmean_worst
-90
-100
-80
-90
-100
-50 -60
-70
-100 -80 -90
-40
-35
-70
-70
-10
0
-30
-10
-20
0
-30
-60
∆ Volmet_all
20
10 0
-10
-20 -30
-40
-80
-70
∆ Volmet_worst
-100
-80
-90
-40
-40
-50
-50
-100
-80
-90
∆ Volmet_worst
30 20 10 0 -10 -20 -30 -40 -50 -60 -70 -80 -90 -100
-80 -78
-90
-100 -100 -100 -100 -100 -100 -100 -100 -100 -100 -100 -100 -100 -99 -99 -98 -98 -98 -97 -97 -96 -93 -93 -89 -88
-100
-10
-3
∆ Volmet_all
-30
-27
1
-70
-64 -61
-70
-64
-50 -46
-60
-52
-23 -12
-26 -20 -4
-40
-40
-37
-20
-18
-60
-100 -100 -100 -100 -100 -100 -100 -100 -99 -98 -98 -98 -98 -98 -97 -96 -93 -93 -87 -86 -85 -81 -78 -73 -69 -64
-50
-45
-100 -100 -100 -100 -100 -100 -100 -100 -100 -100 -100 -100 -99 -99 -96 -96 -95 -95 -93 -92 -89 -85 -84 -84 -79
-47 -46 -45 -43 -43 -43 -38 -34
-50
-48
∆ SUVmean_worst
-30
-40
Figure IV.7. Overview of the different relative parameters of FDG-decrease for the different analyses with corresponding outcome of the patient. No evidence of disease Primary refractory disease Relapse after 12mths Relapse after 23 mths 21
-90
-30
-70
-64
-80
-60
-68 -61 -55 -53
-30
-20
-34 -27
-70
-100
-100 -100 -100 -100 -100 -100 -100 -100 -98 -96 -96 -96 -95 -93 -92 -85 -85 -84 -84 -78 -75
0
0
-10
-20
-100 -100 -100 -100 -100 -100 -100 -100 -100 -100 -100 -100 -99 -98 -98 -98 -98 -98 -97 -96 -95 -93 -93 -90 -87 -80 -78
-60
-90
-51 -47
-50
-80
-66 -63
-40
∆ SUVmean_all
0
-10
-100 -100 -100 -100 -100 -100 -100 -100 -99 -98 -98 -98 -97 -97 -97 -95 -93 -93 -87 -83 -80 -79 -78 -72 -65 -63
∆ SUVmax_all -10
-3 -1
-70
-66 -66 -64 -64 -61 -58 -56 -56
-60
-100 -100 -100 -100 -100 -100 -100 -100 -100 -100 -100 -100 -99 -98 -97 -97 -96 -95 -94 -93 -90 -86 -86 -85 -82
-20
-17 -11
-80
-73
-50
18
10 0 -90
-89
-40
-17
-40 -30
-35
-10 -20
-100 -100 -100 -100 -100 -100 -100 -100 -98 -97 -97 -97 -96 -94 -93 -87 -86 -86 -85 -78 -72 -70 -63 -56 -54 -47
∆ SUVmax_worst -100
45
-30
24
-20
17
0
5
-90
-100
-16 -16 -9
-80
∆ SUVmax_all
-73 -67 -64 -63 -59 -58 -56 -56 -52 -47 -47 -44 -43 -37 -34 -33
-90
-85
-80
-67 -60 -57 -50 -50 -49 -49 -48 -41 -39 -33 -31 -31 -30
-70
-100 -100 -100 -100 -100 -100 -100 -100
-40
-32
-30
-100 -100 -100 -100 -100 -100 -100 -100
-10
-4 -2
-50
-46 -44
-10
-100 -100 -100 -100 -100 -100 -100 -100 -90
-40
-32
-20
4
20
14
-50
-46 -46 -44
-10
-3
-16
-20 -30
-32
-60
-57
-60
-89 -88 -85 -84 -84 -77 -77 -76 -75 -75 -75 -73 -71 -70 -64 -58
0
-67 -57 -52 -50 -49 -49 -49 -46 -40 -31 -30 -29 -26 -22 -11 -11 -10
-46 -44
-64
-70
-85 -84 -80 -77 -77 -76 -75 -75 -75 -74 -73 -71 -70
-100 -100 -100 -100 -100 -100 -100 -100
0
-100 -100 -100 -100 -100 -100 -100 -100 -87
-100
26
-80 -90
1
-70
-46 -46 -44 -34 -32
-50 -60
-100 -100 -100 -100 -100 -100 -100 -100
-100
-100 -100 -100 -100 -100 -100 -100 -100 -90 -87 -84 -82 -80 -77 -76 -75 -75 -73 -73 -70 -66 -66 -64 -58
-100
-100 -100 -100 -100 -100 -100 -100 -100 -87 -84 -78 -77 -76 -75 -75 -73 -73 -71 -66 -64 -62 -62
Early response in DLBCL after 7 days Analysis based on tumor bulk
-50
∆ TLGall
Analysis based on the worst responding lesion 10
-20
0
-50
-60
∆ TLGworst
-10
-20
0
-30
-60
∆ TLGall
∆ TLGworst
Principle of response assessment: influence of different treatments
Numbr of malignant cells
Death
Symptoms ic d rug
wth o r reg d i PET-detection rap
s
(radiotherapy)
chemo
chemo
chemo
chemo
chemo
R/
Diagnosis +chemo
xic to to cy
wth Slow regro chemo
e siv es gr Ag
Cy to s ta t
Time
Intensified therapies are associated with a fast response, but… • . Gallamini A, et al. Haematologica, 2007 – 2xBEACOPPesc – Sens 50%, more false negative lesions – PPV 60%, more false positive lesions
• Avigdor A et al. Haematologica, 2007 – 45 pt advanced staged HL – 2xBEACOPPesc, followed by 4x ABVD – Sens 60% spec 79% NPV 87%, PPV 45% a decrease in accuracy more false negative results more false positive results
Inflammation and its interference with early response assessment Inflammation evoked by therapy more effect of therapy -> more inflammation
FDG-uptake Number of malignant cells
Death
Symptoms
(radiotherapy)
chemo
chemo
chemo
chemo
chemo
chemo
Diagnosis +chemo
PET-detection
Time
Is inflammation important in clinical practice? • High false positive rate after radiotherapy • Jacene et al, JNM 2009, – RIT (Zevalin, Bexxar) – Continuous decrease in FDG-uptake 24 wks after therapy
inflammatory changes with the recruitment of immune cells and high FDG-uptake
High incidence of false positive PET after rituximab in NHL •
Han et al, Ann Oncol 2009 – 51 pt DLBCL+MCL, – Midtherapy (2-4 cycles): PPV= 33%, Sp 68% – Posttherapy: PPV=19%, Sp 80%
•
Haioun et al, Blood 2005 – 90 pt NHL, 37 rituximab, PPV 44%, Sp 70% after 2 cycles
•
Moskowitz et al, JCO 2010, – PET after intensified RCHOP4 in 97 patients – 59 PET neg – consolidation with ICE, excellent prognosis midtreatment negative = excellent prognosis – 38 PET positive – 33 biopsy negative (2 sample error) high frequence of false positive midtreatment, outcome identical as PETnegative patients
False positive PET after rituximab in NHL
Baseline
RCHOP 3
RCHOP 6
inflammatory changes with the recruitment of immune cells ?
Inflammation and its interference with early response assessment Spaepen, EJNM 2003 SCID mice with cyclophosphamide, ex-vivo measurements 100
%
50
viable cells necrosis inflammation ∆ SUV
0 -50
∆ weight -100 0
5
10 time (d)
15
Inflammation and its interference with early response assessment •
Can we improve correlation of FDG-uptake with tumor response? 1. By the administration of steroids? 2. By the use of other PET-tracers: FLT as a marker of cellular proliferation ?
•
Materials and methods – SCID-mouse subcutaneous injected with lymphoma cell line – Treatment with chemo at day 0, half the mice hydrocortisone – Measurements of tracer-uptake by microPET
d0 tumor
d+6
d+9
d+13
d+16
• Does the presence of anti-inflammatory drugs (corticosteroids) influences the FDG-uptake and the cellular respons after chemotherapy? 0
100
SUVmax no cortison
-10
90
SUVmax with cortison -20
80
inflammatory cell fraction no cortison
-30
70
inflammatory cell fraction with cortison
-40
60
-50
50
-60
40
-70
30
-80
20
-90
10
-100
0 0
H&E
CD20
1
CD20
3
4
H&E
H&E
Ki67
2
Ki67
CD20
5
6
7 8 Time (days)
9
10
11
12
13
14
Cyclophosphamide Cyclophosphamide + hydrocortisone Ki67
Brepoels L, Stroobants S, et al. J Nucl Med. 2007
Alternatives for FDG? Proliferation tracers. • Can we improve correlation of tracer uptake with tumor response by using FLT as a marker of cellular proliferation ? – Wagner, Cancer Research 2003 • High uptake in murine model lymphoma, correlation with BrdU in mice • correlation with Ki67 in patients, high grade vs low grade lymphoma
Metabolism of Thymidine
Metabolism of FLT: marker of proliferation TK 1
Cancer: TK1 (x3-x4) degradation ↓
‘metabolic trapping’
Response evaluation by FLT-PET
Rectumca: FDG + FLT before, during and after CRT
FDG d0
FDG post CRT
FLT d0
FLT post CRT
580207v297
Inflammation and early response assessment: is FLT more accurate? Granta cell line (Mantle cell lymphoma) in SCID mouse
A % decrease SUV
60
Cyclophosphamide
40
FDG
*
20 0 -2 0
FLT
-4 0 -6 0 0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
Time (days) FDG and FLT-uptake after cyclophosphamide
Brepoels et al, JNM 2007
Inflammation and early response assessment: is FLT more accurate? A. FDG before treatment
B. FDG after treatment
C. FLT after treatment
Illustration of the high specificity of FLT-PET compared to FDG-PET. (A) PET before therapy shows an extensive lymphoma localization in the proximal tibia (B) After chemotherapy and local radiotherapy, FDG-uptake is still clearly positive but post-radiotherapy changes can not be distinguished from persistent lymphoma (C) FLT-PET after therapy shows a focal uptake in the proximal tibiae which suggests persistent lymphoma (mark the high FDG-uptake in the bone marrow in the non-pathological tibia). The patient relapsed several months later.
Alternatives for FDG? Proliferation tracers. •
Can we improve correlation of tracer uptake with tumor response by using FLT as a marker of cellular proliferation ?
• Metabolism ≠ proliferation: cytostatic and cell cycle targeted agents? Is FLT more accurate in cell cycle targeting therapies?
Inflammation and early response assessment: is FLT more accurate? Mantel cell lymphoma R/mTOR inhibitor
FLT d0
FLT d0 d+7 FDG
Is FLT more accurate in cell cycle targetting drugs? A. FDG
B. FLT
C. FLT
D. FDG
Early response assessment after therapy with mTOR inhibition. (A) FDG-PET/CT before therapy (B) FLT-PET/CT before therapy (C) FLT-PET/CT one week after the first administration and (D) FDG-PET/CT after 6 weeks of therapy The patient obtained a disease free status after a few months of therapy and is still in complete remission (36 months)
Inflammation and early response assessment: is FLT more accurate? Granta cell line (Mantle cell lymphoma)
A
B Cyclophosphamide
% decrease SUV
60
60
40
Temsirolimus
40
*
20
FDG
FDG *
20 0
0
FLT
-20
-2 0
FLT
-40
-4 0
-60 0
1
2
3
4
5
6
7
8
9
10
11
Time (days)
-6 0 0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
FDG and FLTFLT-uptake after temsirolimus Time (days)
FDG and FLT-uptake after cyclophosphamide
12
13
14
Explanation?
histology after mTOR treatment showed a decreased cyclin d1 expression shortly after therapy, which increased again on D+7 Synchronization of the cells? Repair mechanisms? Close interactions of FLT uptake with cellular metabolism
S
TK 1
Other more specific tracers? • Apoptosis: annexin, caspase-3 ([18F]ICMT-11) • Lymphoma specific tracers: Recombinant anti-CD20 antibody fragments,… • 89Zr-Zevalin • Methionine • FET
Opportunities of animal studies • No limitations on numbers of scans, radiation protection: time course of tracer uptake • Standardization • Different treatment regimes, evaluation of the different components of a regimen • Histological confirmation possible, ex vivo measurements of enzymes, …. A. FDG uptake after cyclophosphamide
D0 SUVFDG 3.0
D+1 SUVFDG 2.6
D+7 D+4 D+2 SUVFDG 2.3 SUVFDG 2.1 SUVFDG 3.0
D+9 SUVFDG 2.1
D+11 D+14 SUVFDG 2.0 SUVFDG 4.4
But… • Evaluation of therapy response, not of “sufficient” response. Prognostic significance? • Human cell lines in immunodeficient mice: interference with the immune system? HL? • Syngeneic mice: growth of lymphoma-like pathology, potential to evaluate the effect of new treatment strategies (E.g. vaccination studies, Chaise, 2007, cancer immunol immunother)
• No new more accurate tracers compared to FDG have been developped, potential mainly because of their higher specificity
The future?
Animal studies allow the evaluation of • Interaction of tracers with cellular metabolism • Interaction of therapy with cellular metabolism • Interactions of therapy with uptake of PET tracers