Single Pulmonary Nodule Charecterization: SUVs Are They Trustworthy

1
Single Pulmonary Nodule Charecterization SUVs
Are They Trustworthy?

• Hani A. Nabi, M.D., Ph.D.
• Department of Nuclear Medicine
• School of Medicine and Biomedical Science
• University at Buffalo
• State University of New York
• Buffalo, N.Y. USA

2
PET FDG Imaging in SPN35 of SPN are malignant.
40 of SPN are granuloma.

• Definition
• Parenchymal lesions, well defined less than 4 cm.
• Benign Features Central, concentric or stippled
  calcifications, stable for more than 2 years,
  patients younger than 35 years.

3
PET FDG Features of a Benign SPN

• An SUV of less than 3.5
• SUV Mean Counts/pixel/sec x calibration
  factor
• Injected dose (mCi)/body
  weight (kg).

4
PET FDG Imaging in SPN

• Author Patient Size Sens. Spec.
• (cm)
• Dewan, 1993 30 95 80
• Gupta, 1993 34 0.6-3.0 95 86
• Patz, 1993 51 1.2-6.0 100 89
• MultiCenter 237 -- 96 90

5
Semi-Quantitative Methods

• SUV
• Tumor to background ratios
• SUV versus visual interpretation
• SUV prognostic significance

6
SUV bw mCi/g of tissue mCi injected/patient
body weight (g)SUV bsa mCi/g of tissue mCi
injected/patient BSA in m2
7
Standard Uptake Values of FDG

• Hypothesis
• FDG uptake, normalized to body weight,
  overestimates value in heavy patients.

8
44 patients with malignancies in abdomen or
pelvis, 445-115 kg, studies 60-90 minutes
following FDG injection.Mean SUV bsa normalized
to mean SUV bw. J Nucl Med 94, 35164-167.
9
Results

• Mean SUV bw and bsa same 3.42
• SD of SUV bsa smaller
• Mean SUV bw varies widely with bw, tending to
  overestimate low to intermediate FDG uptake in
  heavy patients and underestimate
  intermediate--higher FDG uptake in light patients

10
SUV, TNT and Visual AnalysisJ Nucl Med 1994,
351771-76

• 107 patients with single pulmonary nodule 67
  malignancies and 27 benign.
• SUV bw were compared to visual interpretation by
  two independent readers.
• FDG uptake grade 1-5 (definitely tumor) based on
  mediastinal uptake (3).

11
Results

• SUV and visual analysis nearly identical.
• No difference between two readers.
• Probability of tumor high (0.93, CI 0.88-1.00)
  when FDG in tumor exceeded mediastinum.

12
Effect of positioning of region of interest (ROI)
in patients undergoing positron emission
tomography (PET) with F-18 FDG and its influence
in varying standardized uptake values (SUV)
13
OBJECTIVES

• EVALUATE SUVS AND HOW THEY ARE AFFECTED WHEN
  ROIS ARE DRAWN IN DIFFERENT PLANES AND IN
  DIFFERENT AREAS OF THE SAME PLANE
• DETERMINE CLINICAL IMPORTANCE OF TUMOR/BKG RATIO
  OF SUVS

14
BEST VISUAL SLICES
15
ROIS ON SAME SLICE
16
THREE ADJACENT PLANES
17
VARIATION OF SUVS OF TUMOR OVER 3 PLANES
18
SUVS OF TUMOR WITHIN ONE PLANE
19
7.88/3.74/13.32
20
(No Transcript)
21
DISCUSSION

• SIGNIFICANT VARIATION
• (p lt 0.009) WAS OBSERVED IN TUMORS OF ADJACENT
  PLANE WHERE SUVS WERE gt20
• MINIMAL VARIATION WAS OBSERVED IN SUVS WHEN
  ROIS OF THE TUMOR WERE DRAWN ON ONE PLANE

22
PROBLEMS WITH SUVS

• UPTAKE TIME
• GLUCOSE LEVEL
• PARTIAL VOLUME EFFECT
• PATIENT SIZE
• ROI EFFECT
• ? NODULE SIZE?

23
SUVs and Nodule Size

• JNM 2002, 43(5) 156, Coronado et al.
• Compared SUVs amongst nodules lt 1.5 cm - gt 3.0 cm
• 45 patients with SPNs
• SUV gt mean
• 21 SPNs 1.5cm or less ( 2.79)
• 18 SPNs 1.6-3.0cm (4.88)
• 7 SPNs 3.0cm or greater (5.16)
• 32/45 FDG positive
• 27/30 proven neoplasia

24

• Nodules 1.5cm or less (21)
• 04cm-1.5cm (mean 0.97cm)
• 11 malignant, 4 FN, 10TN
• FN 1.0cm , FDG but SUVs less 3.5
• Nodules 1.5 cm or less had lower SUVs, a cutoff
  value of 2.0 might be more appropriate.

25
Journal of Hematology Oncology 2008 113
26
Khalaf et al

• Retrospective Study (2 centers)
• Criteria for enrollment
• Positive CT for SPN
• Positive FDG PET
• Histopathological Biopsy
• 173/420 patients fulfilled above criterion

27
(No Transcript)
28
(No Transcript)
29
(No Transcript)
30
Linear regression equation fitted to all nodules
malignant and benign nodules
31
lt 1.0cm
1.1 2.0 cm
32
lt 1.0cm
1.1 2.0 cm
2.1 3.0 cm
gt 3.0 cm
33
Relationship between SUVs and nodules 1.5 cm or
less
SUV max cutoff 2.5 useful in the evaluation of
nodules 1cm or greater, has minimal or no value
in SPN lt 1.0 cm
34
Occurrence of neoplasia in single pulmonary
nodules (SPN) with no visual FDG uptake or FDG
uptake SUVs less than 2.5 established by long
term clinical follow-up.

• Hani A. Nabi, M.D., Ph.D
• L. Li, M.D., Ph.D,
• John Baker, Ph.D
• Department of Nuclear Medicine
• School of Medicine and Biomedical Science
• University at Buffalo
• State University of New York
• Buffalo, N.Y. USA

35
Objective

• Prognostic significance of FDG PET negative
  (visual or SUV less than 2.5) SPN.

36
Patients and Methods

• Retrospective review of all FDG SPN patients
  referred by a single pulmonary group from
  20002004.
• One hundred twenty-eight (128) patients (68
  females, 60 males, mean age 63 years, 56 between
  51-70 years) fulfilled criteria for negative
  scans.
• Mean follow-up 29.3 months (8-52 months).

37
Patients and Methods

• 72 (56) were current or former tobacco users.
• 8 (6.2) occupational exposure.
• 14 (11) history of cancer, other than lung.

38
Results
8
12
20
59
39
Results (no FDG uptake)

• 76 patients (59) had no visual uptake.
• 21 had biopsy.
• 6 (7.9) were malignant.
• 70/76 (92) were negative on clinical follow-up
  of at least 18 months.

40
Results
31.5m
12.3m
Stable
Follow-up
41
Results (SUVs lt 2.5 cm)

• 52 patients (41) had FDG uptake less than 2.5
• 34 had biopsy.
• 23/52 (44) malignant.
• 29/52 were negative for malignancy on follow-up
  of at least 18 months.

42
Results
Nodule types
43
Results
92
55.8
44.2
8
Comparison of benign and malignant nodule number
between FDG uptake and no FDG uptake
44
Benign and malignant SUV uptake
1.84
1.88
Benign Malignant
45
Results

• The difference in proportion in malignancy
  between the no-FDG uptake and FDG uptake groups
  is statistically significant.
• Likelihood ratio Chi-square 23.6.
• P lt 0.001.

46
Conclusion

• SPN with no visual FDG uptake have a relatively
  low likelihood of malignancy (7.9) compared to
  lesions with FDG SUV less than 2.5 (44). This
  study confirmed our previous work (JNM 2006,
  47173) that there is no predetermined fixed
  value of SUV cutoff that is accurately able to
  diagnose malignancy in small pulmonary nodules.