RSNA 2014 

Abstract Archives of the RSNA, 2014


GUS110

Integrated MR/PET in Prostate Cancer Comparison with Conventional Hybrid Imaging

Scientific Posters

Presented on December 1, 2014
Presented as part of GUS-MOA: Genitourinary/Uroradiology Monday Poster Discussions

Participants

Seunghyun Lee, Presenter: Nothing to Disclose
Jeong Yeon Cho MD, Abstract Co-Author: Nothing to Disclose
Sang Youn Kim MD, Abstract Co-Author: Nothing to Disclose
Joongyub Lee, Abstract Co-Author: Nothing to Disclose
Myoung Seok Lee MD, Abstract Co-Author: Nothing to Disclose
Sungmin Woo MD, Abstract Co-Author: Nothing to Disclose
Seung Hyup Kim MD, Abstract Co-Author: Nothing to Disclose

PURPOSE

To evaluate diagnostic value of integrated MR/PET through the comparison of standardized uptake value (SUV) between integrated MR/PET and CT/PET, and the correlation between SUVs of integrated MR/PET and apparent diffusion coefficient (ADC) values of MRI.

METHOD AND MATERIALS

We enrolled 18 patients with histopathologically confirmed primary prostate cancer, who underwent integrated MR/PET using 18F-choline and 18F-fludeoxyglucose (FDG), respectively, and 18F-FDG CT/PET before surgery. The SUV measurements were carried out side by side on corresponding lesions on fused image data sets, and additional ADC measurements also on MR images in MR/PET. Regions of interest were drawn on 12 regions of the prostate based on anatomy, a total of 216 sectors from 18 patients. The SUVs and ADC values from CT/PET and MR/PET were calculated and compared with the receiver operating characteristic curves and the areas under these curves (AUCs) analysis. The comparison of tumor detection rate between SUVs and ADC values of CT/PET and MR/PET were tested with generalized estimating equation (GEE) method.

RESULTS

The average SUV of tumor tissue in 18F-FDG CT/PET (3.09 ± 1.75) was higher than 18F-choline MR/PET (2.75 ± 1.24) or 18F-FDG MR/PET (2.25 ± 1.32). The ROC curve analysis showed that there was no significant better modality for tumor detection, compared with each other (P > .05). The SUV of 18F-choline MR/PET and 18F-FDG MR/PET were associated with tumor detection rate at univariate analysis (P < .05). Multivariate analysis showed that there was 1.63 times more tumor detection in the SUV of 18F-choline MR/PET than ADC value of MRI (P < .0001). The correlation analysis of modalities for tumor detection showed there was significant correlation between SUV of 18F-choline MR/PET and SUV or ADC value of other modalities, including SUV of 18F-choline MR/PET and ADC value of MRI (P < .0001).

CONCLUSION

Our results suggest that there is no evidence of superior modality for tumor detection among MR/PET and CT/PET. There is only statistical significance in SUV of 18F-choline MR/PET, compared with ADC value of MRI.

CLINICAL RELEVANCE/APPLICATION

The new method of 18F-choline MR/PET can be used with confidence in clinical practice combined interpretation of 18F-FDG CT/PET and MRI.

Cite This Abstract

Lee, S, Cho, J, Kim, S, Lee, J, Lee, M, Woo, S, Kim, S, Integrated MR/PET in Prostate Cancer Comparison with Conventional Hybrid Imaging.  Radiological Society of North America 2014 Scientific Assembly and Annual Meeting, - ,Chicago IL. http://archive.rsna.org/2014/14009169.html