Impact of FDG extravasation on SUV measurements in clinical PET/CT. Should we routinely scan the injection site?

Abstract

329

Objectives: Accurate quantification of standardized uptake values (SUV) with 18-F FDG PET is based on the entire injected dose being delivered intravenously. Unfortunately, despite best efforts, extravasation of FDG frequently occurs. Oncology PET scans are routinely acquired with arms raised over the head to reduce attenuation artifacts. Because of this and that most FDG injections involve the arms or the hand, it is common that the injection site is not included in the scan. The purpose of this study was to evaluate the frequency of FDG extravasation, the significance of extravasation on quantification of SUV and the potential benefit of scanning the injection site for clinical patients.

Methods: We evaluated 190 consecutive patients referred for PET/CT (Siemen’s Biograph 16). Routine scans were acquired with an additional 1-minute bed position for the injection site. A large bore IV y-adapter line (22g or larger) was placed and saline allowed to run for 5-10 minutes prior to injection. Injected doses were decay corrected based on calibrated dose (activity + time), time of injection and residual (activity + time). Injection site activity was calculated as: activity in the entire injection site (average activity (MBq/ml) times volume of injection site (cm3)= total activity of injection site. This value was then decay corrected to time of injection. SUVmax for normal tissue (liver, cerebellum, heart and bladder) as well as for metabolically active lesions were calculated based on the original injection dose and the extravasation corrected injection dose. Percent change in both injection dose and resultant SUVmax were calculated for normal tissues and lesions.

Results: Of the 190 PET/CT studies, 39 (21%) had a visible focus of FDG activity at the injection site. Of these 39 patients, 36 had extravasated activity totaling 1% or less of the total injected dose. Three patients (2% of total) had activity greater than 1% of the injected dose (3.0, 7.7 and 17.5%). Normal tissue and lesions demonstrate extravasation corrected SUVmax values that are approximately proportional but slightly greater than the % change in injected dose, depending on the tissue type. Percent change in SUV max ranged from 0-21%

Conclusions: Significant extravasation of FDG occurs rarely when injected through an IV with running saline. When significant extravasation occurs it can have a large impact on SUV values. This could be of critical importance for both diagnostic PET and evaluation of response to therapy. Acquiring a 1-min bed position to include the injection site is critical for accurate interpretation of PET scans in a small number of patients. This may warrant routine acquisition of injection site data. Automated methods for dose correction would be simple and make this clinically practical.

Research Support (if any): None.