Transcript Radiology

Radiology
Scott Schuetze, MD, PhD
University of Michigan
1
Disclosure

I am not a radiologist
2
Utility of imaging
Before diagnosis
 During staging
 During treatment

3
Advanced imaging is
overused prior to referral to
a musculoskeletal oncologist
Drs. Miller, Avedian, Cummings, Balach
Universities of Iowa, Arizona, Stanford,
Connecticut, & Virginia Mason (Seattle)
4
What should the generalist
or primary provider know?

What is appropriate imaging to evaluate
complaint?


Pain
Mass
Imaging for bone vs soft tissue lesion?
 What information is necessary for the
specialist to accept the referral?

5
Study strengths
Geographic diversity – NW, NE,
California, South-west, Mid-west, Tx
 Inclusion of 8 centers
 Prospective consecutive case selection
 Relatively large number of cases
 Pre-defined criteria for imaging utility
 Statistical analysis

6
Study limitations
U.S. perspective
 Pre-defined criteria may be subjective
 Results for bone & soft tissue combined
 Sequencing of imaging not detailed
 Imaging may be driven local radiologists
interpreting radiographs and/or MRI &
recommending additional studies

7
Study findings
MRI useful in majority of soft tissue tumors,
unnecessary in bone lesions
 Geographic differences in CT frequency
(highest in TX, OK, SC)
 Bone scanning and US were infrequent
 PET overused in OK (12%) vs Seattle (0%)


Advanced imaging overused in benign
bone lesions
8
Lessons learned
Orthopaedic surgeons as guilty as primary
care
 CT, bone scans, PET and US are frequently
unnecessary for diagnosis
 Opportunities for regional education?
 Opportunities for education during training?
 Target orthopaedic surgeons, primary care
and/or radiologists?

9
Pre-referral take-away
High quality MRI with contrast for
soft tissue mass
 Plain x-ray for bone lesion
 Let the specialists sort out the rest

10
Pulmonary micronodules do
not impact survival in young
patients
Drs. Gitelis, Cipriano & Kent
Rush Medical College
11
What is the clinical significance
of <1 cm lung nodules on CT?
Patient demographics – age,
occupation, residence, inhalational
substance use/abuse
 Medical history
 Sarcoma sub-type
 Nodule number and distribution
 Slice thickness of scan

12
Which nodules to worry over?
19 yo Ewings
50 yo UPS
35 yo LMS
65 yo liposarc
13
Which nodules to worry over?
19 yo Ewings
?
50 yo UPS
Histoplasma
35 yo LMS
LMS
65 yo liposarc
MALT
14
Study design
96 pt subset of 380 pts
 Age <50 yrs, bone and soft tissue sarcoma
 80% received chemotherapy
 Overall survival endpoint
 4 strata





No lung nodules (47%)
Single nodule <5 mm (26%)
>1 nodule <5 mm (13%)
>1 nodule >5 mm (15%)
15
Study results
75% of nodules <5 mm and 100% of
nodules >5 mm biopsied were sarcoma
 Survival worse for patients with nodule
>5 mm

Figure 1: Influence of Pulmonary Nodules at Presentation on Survival
in Sarcoma Patients Under age 50.
(N = 96, Mean Follow up 47.2 Months)
1
0.9
80%
76%
68%
0.8
Survival
0.7
0.6
0.5
Group 1: No Nodules n = 45
0.4
Group 2: 1 Nodule < 5mm n = 25
0.3
0.2
Group 3: Any number of Nodules < 5mm n = 37
0.1
Group 4: Any number of Nodules > 5mm n = 14
No
difference
between
Groups 1-3
p < 0.05
for Groups
1-3
compared
to Group 4
36%
0
0
10 20 30 40 50 60 70 80 90 100 110
Time from Diagnosis (months)
16
Related studies
71 sarcoma pts <1cm nodules
426 pts with nodules
# of malignant nodules
Patient
group
<5 cm
(n=128)
>5 - <1 cm
(n=118)
>1 - <3 cm
(n=123)
No cancer
13 (32%)
15 (30%)
46 (59%)
Cancer
115 (42%)
103 (69%)
77 (84%)
Pulmonary nodules resected at VATS: etiology in
426 patients
Ginsberg MS, Griff SK, Go BD, et al.
Radiology 213:227-82, 1999
Indeterminate Pulmonary Nodules in Patients with
Sarcoma Affect Survival
Rissing S, Rougraff B, Davis K
Clinical Orthopaedics & Related Research
459:118-121, June 2007
51 pts with osteosarcoma
Nodule size at 1st CT
<5 mm
>5 mm
benign
15 (60%)
7 (27%)
metastasis
10 (40%)
19 (73%)
CT of pulmonary metastases from osteosarcoma:
the less poor technique
Picci P, Vanel D, Briccoli A, et al.
17
Ann Oncol 12:1601-04, 2001
Informal survey results – basis
for nodules as metastatic disease
Ewing AEWS0031: 1 nodule >1 cm or >1
nodules >0.5 cm
 EURAMOS-1: 1 nodule >1 cm or >3 nodules
>0.5 cm
 COG ARST: 1 nodule >1 cm
 French trials: 1 nodule >1 cm
 Italian trials: 1 nodule >0.5 cm
 SARC012: the oncologist, radiologist and
surgeon should use best judgment

18
Conclusions
Metastasis (<5 mm) at diagnosis does not
= poor outcome
 Lung nodules >5 mm should raise
suspicion
 Many lung nodules should raise suspicion
 Standardize criteria for clinical trials?
 Treat for cure!

19
Total lesion glycolysis by
FDG-PET is of predictive
value in soft tissue sarcoma
Drs. Choi, Ha, Cho, Kang, Kim,
Pang & Han
Seoul National University Hospital
20
Potential value of dynamic
imaging in sarcoma
Prognostic information: natural
course of disease
 Predictive information: disease
response to intervention

21
FDG PET semi-quantitative
measurements






SUVmax: maximum pixel value w/i slice with highest FDG
uptake
SUVpeak: average pixel value w/i fixed ROI in area with
highest FDG uptake
SUVave: average pixel value w/i tumor ROI
TBR: average value w/i tumor ROI / average value w/i
blood pool
TLG: SUVave of uptake above minimum threshold x TV
MTV: volume of tumor within ROI in which FDG uptake is
>40% of SUVmax
22
Challenges in
standardization
Dynamic versus static measurements
 FDG administration protocol
 Hardware calibration
 Observer dependent ROI
 Definitions

23
FDG PET and sarcoma
prognosis
Recurrence-free survival
Factor


Disease
HR
95% CI
P value
SUVmax >6
STS
3.2
1.3-8.2
0.015
SUVmax >6
EWS
0.47
SUVmax >6
OS
0.41
SUVmax >15
OS
4.5
1.3-15.3
0.015
SUVmax does not account for tumor heterogeneity
Do other parameters improve prognostic information?
Schuetze S et al. 2005, Cancer 103:339
Hawkins D et al. 2009, Cancer 115:3519
Hawkins D et al. 2005, JCO 23:8828.
Costelloe C et al, 2009, J Nuc Med 50:340.
24
Study data
66 pts with STS
 Retrospective
 AJCC stage






Receiver operating characteristics
I: 16%
II: 24%
III: 46%
IV: 14%
Adjuvant tx


Radiotherapy: 47%
Chemotherapy: 29%
SUVmax: 6
TLG: 250
MTV: 40 cm3
25
K-M analysis of sarcoma
FDG metabolism
P<0.001
P=0.022
P=0.031
Choi E-S et al. 2013, Eur J Nucl Med Mol Imaging DOI 10.1007/s00259-013-2511-y
26
Multivariate analysis factors
affecting PFS
Factor
Value
Univariate
Multivariate
P value
RR
95% CI
P value
AJCC stage
III or IV
0.035
3.36
1.01-11.02
0.047
Metastases
Present
<0.001
5.99
1.81-19.8
0.003
TLG
250
0.001
4.79
1.51-15.23
0.008
SUVmax
6
0.031
0.203
MTV
40 cm3
0.022
0.736
27
Multivariate analysis factors
affecting PFS
Factor
Value
Univariate
Multivariate
P value
RR
95% CI
P value
AJCC stage
III or IV
0.035
3.36
1.01-11.02
0.047
Metastases
Present
<0.001
5.99
1.81-19.8
0.003
TLG
250
0.001
4.79
1.51-15.23
0.008
SUVmax
6
0.031
0.203
MTV
40 cm3
0.022
0.736
Is TLG predictive of response to therapy in
a uniformly treated high-risk population?
28
FDG PET in sarcoma

Potential roles in sarcoma management



Prognostic information / risk of relapse
Predict response to adjuvant therapy
Early marker of response to drug therapy
Foundation of single-institutional experience
(variability minimized)
 Need more experience in multi-site trials
(more variability)

29
Thanks to the presenters &
session chairs
30