Login to Access Video or Poster Abstract: MP80-09
Sources of Funding: none

Introduction

FDG-PET imaging is a valuable tool to guide additional treatment recommendations in the post-chemotherapy setting for metastatic seminoma. However, false positive results can be a problem. We sought to identify a new methodology to interpret FDG-PET scans using a more objective approach to reduce false positive results.

Methods

We identified patients who had FDG-PET imaging available for re-review with a diagnosis of germ cell tumor at our institution from 2006 to 2016. Twenty-six scans were identified. All images were re-reviewed by an experienced radiologist who was blinded to patient treatments and outcomes. Radiographic variables recorded were mass size, standard uptake values (SUV), liver and blood pool values, and date of scan. Liver and blood ratios were calculated for each scan by dividing the SUV of the index lesion by the liver and blood pool values, respectively. A ratio of ≤1 would be considered a negative scan. A 5-point scale was assigned to each scan based on the dominant FDG-avid lesion using a similar system to the Deauville scale for lymphoma with 5 representing significant uptake and 1 for no uptake.

Results

A total of 26 patients were identified. The median follow-up from the PET scan was 21 months (range 1-96). The sensitivity, specificity, positive predictive value (PPV), and negative predictive value (NPV) of the original PET scan interpretation was 100%, 81%, 77%, and 100% respectively. If the liver ratio was included as an objective measurement, the sensitivity, specificity, PPV, and NPV would have improved to 100%, 88%, 83%, 100%, respectively. Thus, the increase in specificity would have resulted in a decrease of false positive results. Of the 26 PET studies, 3 (12%) were false positives and 0 (0%) were false negatives. Four patients underwent a RPLND due to positive findings on the PET study. Of these, 3 were found to have seminoma and 1 had necrosis on final pathology. The median SUV value of the 3 PCRPLND patients with seminoma was 6.7, with a liver ratio of 2.68. The patient with necrosis had an SUV of 2.5, with liver ratio of 0.9. The blood ratio and 5-point scoring system did not add additional significant information.

Conclusions

By including the liver ratio in interpreting PET scans, we believe we can reduce the number of false positive scans.

Funding

none