PET/MR Characterization of Renal Cell Carcinomas

Simultaneous PET/MR imaging [1]-[7] offers exciting opportunities to visualize and quantify soft-tissue tumors [2]. MRI offers superb soft-tissue contrast for anatomical information as well as a flexible suite of other techniques providing functional and physiological information. PET offers sensitive molecular imaging via radioactive tracers and is widely used for assessment of tumor glucose metabolism. Together, the two modalities provide complementary, synergistic information. Because of MRI's superior soft-tissue contrast, it is considered a much better anatomical guide for PET quantitative analyses for tumors in soft tissue regions as compared to standard-of-care PET-CT. Simultaneous PET-MR also provides inherently-aligned PET and MR, efficient simultaneous acquisition, and the opportunity for new approaches to PET quantitative analysis guided by detailed MR images.

Nonmetastatic clear cell renal cell carcinomas are generally treated with nephrectomy, although research has subclassified them into two [8] or as many as four [9] subtypes based on gene expression; the subtypes have been shown to have differing prognoses and progression [8]-[11]. The investigators focus primarily on the two subtypes ccA and ccB [8] and ask whether these two exhibit distinct PET-MRI imaging characteristics that can be used to subtype them noninvasively. An eight-subject pilot study conducted on PET-MRI at UNC demonstrated that quantitative PET-MRI imaging, using nonstandard measures incorporating the entire tumor, was strongly correlated with presence or absence of ccB patterns in the tumors [12]. Transcriptome analysis results from the study justified the import of PET in noting that many of the upregulated genes associated with ccB were involved in glucose transport and metabolism, the mediators of 18-FDG uptake. Thus, there is a strong biologic rationale as to why PET-MRI could predict ccB burden, but the small number of subjects in that study limits confidence.

Also, the fact that these tumors are highly heterogeneous suggests that imaging alone provides an opportunity for early characterization of ccB burden across the entire tumor where gene expression profiling would be invasive and costly. Also, with the excellent soft-tissue contrast of MRI, the system has the potential for greater anatomical detail to guide interpretation of FDG activity. UNC has one of the few PET-MR scanners in the country, making it one of few centers in the US capable of conducting this study.

This is a prospective study of the use of combined PET-MR [1-7] for prediction of tumor subtype composition in ccRCC. Additional data analysis using the prior study from UNC suggests a strong correlation between the fraction of the tumor exhibiting moderate to high FDG uptake and the number of cores found to be ccB type. These results guide the hypothesis of this study: the metabolically-active tumor volume fraction derived from simultaneous PET-MRI imaging of clear-cell renal cell carcinomas is correlated with the fraction of tumor cores classified as ccB subtype from transcriptome analysis. If so, in the future, application of PET-MRI may offer prognostic information and opportunities for risk stratification.