2013. It Takes Two to Tango: Detection Efficiency of 18F-FDG PET-CT and High-Resolution Contrast Enhanced CT in the Evaluation of Hidden Primary
Authors* Denotes Presenting Author
Maimoona Siddique *;
Pakistan Kidney and Liver Institute & Research Centre
LINCS Diagnostics & IR Services
18F-FDG PET-CT is an astute detective in unraveling unknown primaries in cases presenting either with metastatic disease or constitutional symptoms that remains unspecified despite thorough diagnostic evaluation. We took a slingshot by adding anatomical probe high-resolution contrast-enhanced (HR CECT) in the diagnostic regimen.
Materials and Methods:
Retrospectively, age and sex-matched cases with unknown primaries or negative nodal or extra-nodal diseases who underwent multimodal imaging (combined 18F-FDG PET-CT and HR CECT) from June 2019 till July 2020 were evaluated. We assessed the combined work of both as compared with them working alone. Tissue diagnosis and follow up particularly in negative cases were considered the reference standard.
Statistical analysis included detection rate, sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV), and area under the curve (AUC) for FDG PET alone and combined molecular-diagnostic imaging. Linear regression was calculated to interpret PET-CT outcome based on background-corrected SUVmax values.
Of 124 cases [66 males, 58 females; age 18 – 73 y, mean: 49.8 y], pre-scan biopsy was suggestive of metastases in 28% (n=35), inconclusive in 14% (n=17) and rest of 58% (n=72) had only constitutional symptoms.
89 patients had hypermetabolic nodal and/or extra-nodal disease on PET-CT scans; 52% (64/124) turned out to be true positive for a primary tumor on biopsy and false positive in 20% (25/124) cases due to underlying granulomatous disease. Cases with FDG non-avid disease, 3%(4/124) were false negative, while the rest of 25% (31/124) were true negative. The primary tumor was detected in gastrointestinal (n=14), pancreaticobiliary (n=8), genitourinary (n=4) and respiratory (n=6) systems. Four cases had breast carcinoma, 24 of lymphoproliferative origin, three of malignant melanoma, and miscellaneous sites in 3 cases. Putting all the pieces of jigsaw puzzle revealed that combined molecular – high-resolution anatomical imaging was more precise in picking the true positives.
Detection rate, sensitivity, specificity, PPV and NPV for FDG-PET alone were 77%, 94%, 56%, and 76%, 73% and 88.5% respectively and for combined PET – high resolution CECT were 91%, 98%, 83%, 88% and 97% respectively. The receiver operating character curve (ROC) revealed that the working model of two was accurate in detecting unknown primary - AUC 0.978; SE=0.018; 95% CI: 0.943-1.000 versus AUC 0.751; SE=0.541; 95% CI: 0.645-0.857 for PET alone. There was statistically significant (P-value <0.001) difference in diagnostic efficiency of PET imaging by addition of HR-CECT scan.
High-resolution CECT when added to 18F-FDG PET-CT works much better providing higher detection quotients and reducing the number of false positives. In unison, both work as an effective tool for identifying occult primary pathology.